The Denmark data breach has exposed the names, addresses and CPR numbers of approximately 8.8 million people after unauthorized users exploited a Danish company's legitimate access to the country's Central Person Register (CPR), according to the Danish government. The CPR administration became aware of irregular behavior in the system on October 2, after identifying activity that had occurred during September. An investigation found that unauthorized individuals had used a private Danish company's legal access to search the CPR system and obtain information on millions of registered people. The CPR currently contains information on approximately 11 million registered people, including those who are living in Denmark, people who have moved abroad and deceased individuals.

Denmark Data Breach Involved Legitimate CPR Access

The Denmark data breach did not result from unauthorized access being granted directly to the attackers. Instead, unauthorized users abused a private company's legitimate access to the CPR system. Under Section 38 of Denmark's Civil Registration Act, private companies with a legitimate interest can receive information from the CPR about specific groups of people they have individually identified in advance. Companies must also be entitled to receive the information under the General Data Protection Regulation and Denmark's Data Protection Act. The investigation found that unauthorized individuals used this legitimate access to retrieve information from the register. The exposed information includes names, addresses and CPR numbers, which serve as personal identification numbers in Denmark. The CPR administration said the unauthorized access did not include the names and addresses of people who had registered for name and address protection.

Investigation Into CPR Data Breach Continues

The CPR administration has stopped the company's access to the system and is working with specialists and relevant authorities to establish how the incident occurred. The case has been reported to the Danish Data Protection Authority, while police are investigating in cooperation with other relevant authorities. Authorities said the investigation remains in its early stages, and it is not yet possible to identify who was responsible for the unauthorized access. The CPR administration also warned that further investigation could lead to changes or clarification of the information currently available about the incident. Denmark's Minister of Research, Education and Digitalisation Christina Egelund described the incident as serious and said a thorough security review of the CPR system would be conducted. Additional measures have also been launched to prevent similar incidents.

What the Denmark Data Breach Means for Citizens

Authorities have reminded citizens not to provide passwords or other confidential information during phone calls or email exchanges, even when the person making contact appears to know their name, address or CPR number. Guidance is available through Denmark's Sikker Digital service, while the Cyberhotline for digital security has extended its opening hours from 8 a.m. to midnight in the coming days. The incident also highlights the risks associated with legitimate access to sensitive databases. In this case, the access used to obtain the information was originally available to a private company under Denmark's existing rules for accessing CPR data.

Cybersecurity Awareness Month Puts Data Protection in Focus

The Denmark data breach comes during October's Cybersecurity Awareness Month, a period focused on improving awareness of digital security risks. The incident highlights why protecting access to sensitive personal information remains an important part of cybersecurity. For organizations handling large volumes of personal data, the case also shows the importance of monitoring legitimate access and identifying unusual activity. However, Danish authorities have stressed that the investigation is still underway and that the full circumstances of the incident have yet to be established.

OpenAI rogue AI agents carried out unauthorized activity across Wikimedia platforms, including wiki edits, attempts to use a public note-taking service as a proxy and millions of automated requests, according to an investigation by the Wikimedia Foundation. The Foundation said it identified activity it believes was conducted by agents operated by OpenAI after investigating reports that so-called “rogue” AI agents had attempted to break into websites and online services. Wikimedia said it found no evidence that its systems or data were compromised or that its platforms were used to coordinate agents. However, the investigation identified several forms of unauthorized activity involving AI agents.

OpenAI Rogue AI Agents Made Unauthorized Wiki Edits

Wikimedia identified edits that it believes came from OpenAI rogue AI agents. Almost all of the edits were made in wiki sandbox areas and were not published on pages visible to general readers. The activity also included several edits involving the configuration of a citation tool. Wikimedia said these edits appeared potentially malicious and may have been intended to misuse the tool as a proxy for fetching data from remote services. Wikipedia allows bots to make edits when they are disclosed and approved by community editors. Wikimedia said no such approvals were sought in these incidents. The Foundation said its investigation did not find evidence that the edits resulted in a compromise of Wikimedia systems or data.

Etherpad Attempts and Agent Activity

Wikimedia also found unsuccessful attempts involving Etherpad, a public note-taking tool operated by the Foundation as a community service. According to Wikimedia, agents believed to be operated by OpenAI attempted to use Etherpad to fetch data from other websites as a proxy. Other agents also appeared to use the service to take notes about their tasks, although Wikimedia said this did not appear to develop into coordination. The findings come amid reports of AI agents using public websites and online services in unintended ways. Wikimedia said OpenAI agents have previously been observed using public platforms as communication channels. In one previously reported incident, agents repurposed a wiki website as a message board to exchange information about how to circumvent evaluation tasks. Similar behavior was reported in the Hugging Face incident, where agents used an OpenAI file-sharing service to coordinate activity during a cybersecurity assessment.

Millions of Requests Hit Wikimedia Infrastructure

The investigation also found significant automated traffic associated with agents believed to be operated by OpenAI. Wikimedia said the agents made millions of automated requests to its public APIs, crawled millions of pages, primarily from Wikidata and Wikimedia Commons, and generated hundreds of thousands of queries to the Wikidata Query Service. The Foundation said the traffic may have contributed to a partial outage affecting the Wikidata Query Service in May. The scale of the activity adds to existing concerns over automated traffic on Wikimedia projects. The Foundation previously reported that bandwidth usage had increased by 50% because of increased bot activity since 2024, while 65% of the most resource-consuming traffic on its projects was coming from bots.

OpenAI Calls Wiki Incident a Misalignment Issue

OpenAI confirmed the incident after Reuters first reported it. The company described the “wiki incident” as an example of misalignment, referring to situations where an AI system does not follow human intentions. OpenAI also argued that the AI industry does not have a common standard for disclosing incidents involving models behaving in unintended ways. The company denied claims that its lawyers had pressured employees to keep the incident quiet. Wikimedia said the incident highlights the growing challenges created by AI agents operating across open online services. The Foundation said it found no evidence that its platforms were used for agent coordination, but warned that the activity placed additional pressure on infrastructure maintained for public use. OpenAI CEO Sam Altman also said in a Monday interview with Politico that society should accept some negative consequences in exchange for the benefits of AI technology. [caption id="attachment_114596" align="aligncenter" width="600"] Image Source: X[/caption]

The Wikimedia investigation adds to recent reports of AI agents interacting with public websites and online services in unintended ways. The Foundation said such activity can create challenges for website operators, particularly when agents generate large volumes of automated traffic or attempt to use public services for purposes beyond their intended function.

Authored By Salleh Kodri, Regional Director, Solutions Engineering, ASEAN, Cyble

For the last decade, cybersecurity has spent a lot of energy trying to solve a visibility problem. We wanted more logs. More telemetry. More threat intelligence. More detection rules. More dashboards. And to be fair, we got them. Today, many organisations have SIEM, EDR, NDR, attack surface management, identity security, vulnerability management, cloud security and multiple threat intelligence sources. The average SOC can probably see more of its environment than at any point in the past. Yet incidents still happen where, afterwards, somebody inevitably asks: "Did we already have the indicators?" Quite often, the answer is yes. That tells me the next cybersecurity problem is not simply visibility. It is decision latency. Seeing something is one thing. Understanding what it means, deciding what matters and taking the right action before the situation changes is something else entirely. And Agentic AI is about to make that gap much more obvious.

We Are Crossing a Speed Threshold

There is a lot of discussion right now about AI-powered cyberattacks. I think we need to be careful with the hype. AI does not suddenly make every traditional attack technique obsolete, nor does it mean tomorrow's attacker will be an entirely autonomous machine. The more immediate change is simpler. AI compresses time. Reconnaissance can become faster. Target discovery can become faster. Vulnerability analysis can become faster. Social engineering can become personalised at a scale that was previously difficult. Infrastructure can be analysed, changed and replaced faster. And increasingly, AI systems can help determine what to do next. Agentic systems take this one step further because they are designed around a loop:

Observe -> Reason -> Decide -> Execute -> Reassess

Potentially with less human involvement at every stage. So when I think about Agentic AI in cybersecurity, I do not start with the question, "How intelligent will the AI become?" I start with another question: What happens when the attacker can make decisions faster than the defender? That, to me, is the more immediate challenge.

The SOC Does Not Have an Alert Problem. It Has a Decision Problem

In conversations with security teams across ASEAN, I rarely hear anyone asking for fewer sources of security data. They already have plenty. The problem is making sense of all of it quickly enough. Imagine an incident where an organisation detects suspicious authentication activity. The identity platform sees one piece. The endpoint platform sees another. Threat intelligence knows something about the IP address. Attack surface management identifies an exposed service. The vulnerability platform knows there is a weakness on that asset. Digital risk monitoring may already have discovered credentials belonging to the same employee somewhere outside the organisation. Every tool may technically be doing its job. But somebody still has to connect the dots. Is this actually malicious? Is the account privileged? Is the asset internet-facing? Is the vulnerability exploitable? Are the credentials genuine? Has the infrastructure been associated with a known threat actor? Should we isolate the endpoint? Should we disable the account? What happens to the business if we do? This is why I believe cybersecurity increasingly suffers from decision latency. For years, security architecture has been optimised heavily around detection latency. How quickly can I see something? Now the question has to become: How quickly can I understand enough of the situation to make the right decision? If five security systems generate five different pieces of the same attack, but an analyst has to manually correlate them before anybody understands the risk, we are still operating at human speed. That becomes a bigger problem when the attacker is not.

Threat Intelligence Has to Move Closer to Action

I have spent a significant part of my career working around threat intelligence, and one thing I believe strongly is that intelligence has to change a decision. Otherwise, it risks becoming another source of information. Traditionally, threat intelligence has often followed a model like this:

Collect -> Analyse -> Report -> Distribute

There is still an important place for that, especially for strategic intelligence. But operational security requires something closer to:

Detect -> Enrich -> Correlate -> Understand -> Decide -> Act

Take something as simple as a leaked credential. A username and password belonging to an employee appears in an underground marketplace. On its own, that is intelligence. Useful, but incomplete. Now imagine immediately correlating that information with the person's identity, privilege level, external applications, exposed infrastructure, endpoint activity, known vulnerabilities, suspicious infrastructure and current adversary behaviour. The question is no longer: "Did we find a leaked credential?" It becomes: "Does this credential create a credible path into something important, and do we need to act now?" That is a very different security outcome. And this is where I see the real potential of AI in the SOC. Not another AI assistant creating another summary. Not another chatbot sitting on top of another dashboard. AI should reduce the distance between intelligence and action.

ASEAN Has its Own Reality

This conversation becomes particularly interesting when viewed from ASEAN. One thing I have learned working across the region is that there really is no single "ASEAN cybersecurity maturity level." Singapore is different from Malaysia. Malaysia is different from Indonesia. Indonesia is different from Thailand or the Philippines. And even within each country, a major bank, a government agency, a telecommunications provider, a manufacturing company and a mid-sized enterprise can have completely different levels of cybersecurity maturity. Some organisations have mature SOC operations with substantial automation. Others are still strengthening basic identity security, vulnerability management, endpoint visibility and incident response. Yet they may ultimately face the same AI-assisted adversaries. That creates an uncomfortable reality: ASEAN may face machine-speed threats while many organisations are still operating human-speed security processes. We cannot solve that simply by telling everyone to hire more SOC analysts. The talent challenge is already real. Instead, I see AI becoming a force multiplier for analysts - handling repetitive enrichment, correlation, evidence collection and initial investigation so people can spend more time on the decisions where judgement actually matters. But this does not mean automating everything. In fact, the more powerful our automation becomes, the more careful we need to be.

Where AI Should Not Act Matters Too

This is particularly important for critical infrastructure. ASEAN economies depend heavily on energy, manufacturing, transportation, telecommunications and other essential services. Many of those environments include operational technology and legacy systems where the consequences of getting an automated decision wrong can be very different from traditional IT. Automatically isolating an employee laptop suspected of compromise is one thing. Automatically interfering with a production environment, operational control system or critical service is another. The cybersecurity decision may be technically correct and operationally disastrous. That is why I keep coming back to this principle: The faster AI can act, the more important it becomes to understand where it should not act. Context matters. Business impact matters. Safety matters. Authority matters. That is also why I do not believe the future SOC is simply an autonomous SOC. It has to be a governed autonomous SOC.

Agentic AI Needs a Trust Architecture

It would be dangerous to interpret Agentic AI as simply giving an AI agent access to every security tool and saying, "Go and defend us." Trust has to be engineered into the architecture. I think about this in three practical levels. Observe. Let the AI collect information, correlate events, investigate and build context. Humans retain the decision authority. Assist. Let the AI recommend an action, prepare the workflow and present the evidence. Humans approve significant actions. Act. Allow the AI to execute predefined actions automatically - but only within clearly established policies, authority boundaries and confidence thresholds. Most organisations will probably use all three at the same time. A low-risk containment action might be automated. Disabling a privileged identity may require approval. Taking action against a production system should probably have an even higher threshold. The technology is actually the easier part. The harder questions are governance questions. What agent is acting? What authority does it have? What evidence informed the decision? What is it prohibited from doing? When must it stop? When must a human become involved? And afterwards, can we explain exactly what happened and why? Autonomy without accountability is not resilience. It is another form of risk.

'Human in the Loop' is Not Enough

I also think we need to challenge one phrase that gets used constantly in AI discussions: Human in the loop. It sounds safe. But what does it actually mean? If an AI system generates 50 recommendations during an attack and every one of them requires an analyst to click "Approve," technically we still have a human in the loop. We have also recreated the bottleneck we were trying to remove. The better model may increasingly be human on the loop rather than human in every loop. Humans establish the rules. Humans define the authority. Humans establish thresholds, escalation conditions, prohibited actions and risk appetite. Machines operate inside those boundaries. And when the uncertainty, context or consequence exceeds those boundaries, the system escalates to a person. Machine-speed defence does not require removing humans from cybersecurity. It requires being far more precise about which decisions actually need a human.

The SOC Needs to Become a Decision System

This changes how I think about the future SOC. The traditional SOC workflow is roughly:

Alert -> Queue -> Analyst -> Investigation -> Response

The emerging model looks more like:

Signal -> Intelligence -> Context -> AI reasoning -> Decision -> Controlled action

That does not make the analyst less important. I think it makes good analysts more important. They should spend less time copying indicators between systems and more time understanding adversary behaviour, investigating complex incidents, assessing business impact, making judgement calls and defining how autonomous systems should operate. AI should automate the work around the analyst, not automate away the analyst's judgement. Because ultimately cybersecurity decisions are rarely purely technical. Blocking something can interrupt a business process. Disabling an identity affects a person. Taking a system offline can affect customers. Changing something in operational technology can create physical consequences. That context is what separates security automation from responsible security automation.

ASEAN can Leapfrog

There is also an opportunity here for ASEAN. We do not necessarily have to replicate every generation of security architecture built elsewhere. Some organisations in the region may be able to move directly from fragmented security operations toward more integrated, intelligence-driven models. But we cannot assume AI will compensate for weak foundations. Identity still matters. Asset visibility still matters. Security hygiene still matters. Threat intelligence still matters. Incident response still matters. And regional collaboration increasingly matters. As our economies become more connected, cyber risk does not respect national boundaries. A compromised supplier in one ASEAN country can affect customers in another. A threat actor targeting financial institutions may operate across several markets simultaneously. Infrastructure, cloud platforms and digital ecosystems increasingly cross borders. That means strengthening ASEAN's cyber resilience will require more than technology. It will require information sharing, public-private cooperation, critical infrastructure resilience, cross-border incident coordination, talent development and sensible AI governance. This is why the conversation taking place around Singapore International Cyber Week and GovWare is timely. SICW's discussion around shaping the new cyber order, and GovWare's focus on cyber at agentic velocity, are ultimately asking variations of the same question: How do we move faster without losing control?

Trust Becomes the Control Plane

I do not think the next cybersecurity race is simply going to be: AI attacker versus AI defender. That is too simplistic. The real advantage will belong to whichever side can understand what is happening, make the right decision and act faster. But for defenders, there is an additional requirement. We must do it without losing control of the decision itself. Speed will become a security capability. Intelligence will provide context. AI will provide scale. But trust must determine what happens next. As cybersecurity moves closer to machine speed, our security architecture will have to move with it. That does not mean every decision should become autonomous. And faster certainly does not automatically mean safer. The organisations that get this right will be the ones that combine machine-speed intelligence, machine-assisted decision-making and human accountability. For me, the formula is relatively straightforward:

INTELLIGENCE + CONTEXT + AI + GOVERNANCE -> TRUSTED ACTION

Everything before the arrow helps us move faster. Everything around it determines whether we should trust the outcome. And that is ultimately what trust at machine speed has to mean.

Google has suspended its Open Source Software Vulnerability Rewards Program (OSS VRP), a vulnerability search initiative that paid researchers for finding flaws in the company's open-source software. The pause took effect on October 1, 2026. Google blamed a "significant rise" in automated and AI-generated reports, most of which turned out to be invalid.  Google announced the suspension in posts on X and on the program's website. It said it would provide "an update" in the first quarter of 2027. According to Tom's Hardware, Google engineers and open-source maintainers had been overwhelmed by reports that were invalid or contained hallucinations. 

Why Google Halted OSS VRP 

"This pause is due to a significant rise in automated submissions, the vast majority of which are not valid," the company said. In the meantime, Google is encouraging participants to look at its other bug bounty programs.  As of October 1, 2026, the OSS VRP no longer accepts product vulnerability reports. Some reports involving Google Cloud repositories that affect Cloud products may still be accepted through the Cloud VRP. Flaws in open-source projects closely tied to Cloud or AI products can go to the Cloud VRP or the AI VRP. Researchers can also turn to the Patch Rewards Program, which pays for security improvements to Google's open-source projects. Reports submitted before October 1 are not affected.  Product vulnerabilities previously covered flaws that substantially affected the confidentiality or integrity of user data. Examples included memory corruption in file format parsers or network protocols, sanitizer failures, path traversal, and insecure defaults or code examples in documentation.   For OT0 and OT1 repositories, memory corruption reports required exact OSS-Fuzz reproduction steps or an already merged patch. Non-memory corruption reports did not need a patch. Google also urged researchers to build OSS-Fuzz integrations for important projects that lacked them before reporting. 

How Project Tiers Shape Vulnerability Search Rewards 

Repositories fall into four tiers: 

• Flagship (OT0) projects are the most sensitive and carry the highest payouts. 
• Important (OT1) projects have significant community impact or security criticality. Both OT0 and OT1 are listed at github.com/google/bughunters. 
• Standard (OT2) covers active repositories with stable packages on registries such as npm or PyPI. There is no published list for this tier. 
• Low-priority (OT3) includes experimental, sample, research, and archived projects. These earn no financial rewards. 

Supply chain compromises pay $3,133.7 to $31,337 for OT0, $1,337 to $13,337 for OT1, and $500 to $3,133.7 for OT2. Other security issues, such as leaked write-access credentials or weak passwords on third-party CI systems, earn $1,000 for OT0 and $500 for OT1. OT2 and OT3 projects receive nothing for product vulnerabilities or other security issues.  A reward panel sets final amounts based on security impact alone. It may pay more for clever or wide-reaching bugs and less for those that depend on hypothetical flaws. It typically pays once per root cause, though bonuses of about $1,000 are possible for especially clever findings or well-written reports. Researchers can donate their rewards to charity. Rewards left unclaimed after 12 months go to a charity of Google's choosing.  Several categories are excluded: 

• Issues rooted in downstream integration 
• Typosquatting without proof that build artifacts were compromised 
• Insecure installation instructions 
• Social engineering 
• Bugs with negligible impact 

Google also asks researchers to avoid DoS attacks, black hat SEO, spam, and high-traffic automated testing tools. Reports go through Google's vulnerability form and should include a buildable proof of concept, the affected versions, an impact description, and an attack scenario. 

Nihon Keizai Shimbun disclosed on October 4 that a Nikkei cyberattack had compromised employees' Microsoft 365 business software accounts. Attackers then used those accounts to send roughly 9,000 spoofed emails to people inside and outside the company.   The publisher also revealed a second, separate breach: a cloud service it uses had been accessed without authorization since late July, possibly exposing the personal information of 1,646 employees and business partners.  In a statement, the company said: "We deeply apologize for the inconvenience and concern caused to all those affected. We ask everyone to remain vigilant against suspicious emails while we continue our investigation. We take this incident seriously and will further strengthen our security measures to ensure thorough information management." 

How the Nikkei Cyberattack Unfolded?

A third party gained unauthorized access to the Microsoft 365 accounts that employees use for email and other work functions. On September 30, the attackers used the hijacked accounts to send about 9,000 emails containing links to malicious websites. The recipients included Nikkei staff as well as numerous external sources and contacts who had previously corresponded with employees.  Because the messages came from Nikkei's genuine accounts, the sender address alone gave recipients little reason for suspicion. The attack is believed to have exposed recipients' names and email addresses, along with the content of some messages.  Nikkei has since changed the affected passwords and says it has detected no further unauthorized logins. It has contacted each recipient individually and asked them to delete the emails. 

Google Workspace Breach Dates Back to July 

In a separate incident, outsiders also logged in to the Google Workspace cloud accounts that employees use for work. Since late July, names, email addresses and other information belonging to 1,646 people may have leaked. Nikkei learned of the problem in early August, when Google notified the company. According to Nikkei, the leaked data did not include information on readers or sources, and no secondary damage has been confirmed.  The incidents unfolded in this order: 

• Late July 2026: Unauthorized access to Google Workspace begins. 
• Early August 2026: Nikkei discovers the access after Google notifies it. 
• September 30, 2026: Attackers hijack Microsoft 365 accounts and send about 9,000 spoofed emails. 
• October 4, 2026: Nikkei announces the incidents publicly and reports them to the Personal Information Protection Commission. 

Nikkei Cyberattack Reported to Regulator 

Nikkei has reported both incidents to Japan's Personal Information Protection Commission. Under the Act on the Protection of Personal Information, organizations must report to the Commission and notify affected individuals when a breach caused by unauthorized access could harm people's rights and interests. Breaches affecting more than 1,000 people are also subject to mandatory reporting.   A preliminary report is due within roughly three to five days of discovery. A final report is due within 30 days, or within 60 days if the breach was carried out for malicious purposes. Nikkei is still investigating the full scope of the breach and the number of personal records involved.  The company warned that spoofed emails posing as Nikkei or group company staff may increase. It asked anyone who receives a suspicious message to contact it through its inquiry form. 

The Shinhan Bank cyberattack has exposed personal and financial information belonging to about 25,000 customers in South Korea, with investigators examining whether advanced artificial intelligence tools were used to carry out the intrusion. The leaked information reportedly includes customers’ names, phone numbers, loan borrowing details and annual income. South Korea’s Financial Supervisory Service has launched an emergency on-site inspection as authorities investigate how the breach occurred and whether the suspected use of AI points to a wider risk for financial institutions.

Shinhan Bank Cyberattack Under Investigation

Shinhan Bank said the information was leaked on Wednesday after hackers gained access to its systems. The bank immediately formed a response team and took measures aimed at limiting further damage. These measures included blocking access from external IP addresses and suspending the affected services. The bank said it had completed safety measures related to the incident. Sources cited in the reports said the attackers are suspected to be based overseas and may have used advanced AI tools during the attack. Experts said the bank may not have been specifically targeted, but could instead have been caught in attacks against vulnerable online platforms. The Financial Supervisory Service sent a team to the bank to examine the incident. Its investigation is expected to take months as officials determine the nature and extent of the data exposure.

AI Agent-Assisted Attacks Raise Concern

The suspected use of AI has added another concern to the Shinhan Bank breach. Cybersecurity experts said AI tools originally developed or shared for defensive purposes can also be adapted for malicious activity. Mun Chong-hyun, director at cybersecurity firm Genians, said several recent attacks in South Korea have involved such AI tools. He described the technology as a double-edged sword because capabilities created for cybersecurity can also be used to facilitate cybercrime. The concern is not limited to Shinhan Bank. An industry official said the wider financial sector could be exposed to AI agent-assisted attacks, and that financial institutions should conduct their own security checks to identify potential weaknesses and abnormal access attempts. The incident also highlights the type of information that can be exposed in attacks against banks. Unlike a breach involving only basic contact information, the Shinhan incident reportedly exposed both personal details and financial information.

South Korean Banks Face More Data Breaches

The Shinhan Bank cyberattack comes amid several recent incidents affecting South Korean lenders. KB Kookmin Bank said last Friday that personal information belonging to 119 customers had been leaked following an external intrusion. Hana Bank also reported a separate hack affecting 89 customers. South Korea’s Financial Services Commission has already held a meeting with local banks over the recent breaches and is expected to hold another meeting this week. The latest incidents have raised concerns about how AI-powered hacking could affect the financial sector, particularly if attackers can use automated tools to identify vulnerable systems or support attacks at greater scale. The Shinhan incident is relatively small compared with some of South Korea’s largest data breaches in terms of the number of affected people. However, the combination of personal and financial information makes the breach significant. Sungho Hwang, Korea country manager at NordVPN, said the exposed information could be used to create personalized scams. He also noted that generative AI has made such scams more convincing.

South Korea Faces Wider Cybersecurity Pressure

The banking incidents come as South Korea cyberattacks have increased across other sectors. Cyberattacks targeting South Korean military systems reached 18,951 cases in 2025, the highest annual figure in five years. The number was up 31 percent from 2024 and 108 percent from 2022. South Korea has also expanded its use of AI for defensive cybersecurity. The Ministry of Education and the Korea Education and Research Information Service operate an AI-based system designed to detect suspicious activity and issue automated alerts across 435 monitored institutions. In 2025, the system detected about 480 million potential cyber threat indicators, of which around 86,000 were confirmed as actual intrusions. As the investigation into the Shinhan Bank breach continues, the suspected role of AI is adding to concerns that financial institutions may need to account for increasingly automated forms of cyberattacks.

Japan is stepping up its plans to find cyber attackers who may already be inside the networks that keep essential services running. Under a new fiscal 2027 initiative, the government plans to work with private companies to strengthen Japan threat hunting across critical infrastructure, including power utilities and telecommunications operators. The focus is not only on stopping attacks after they are detected, but on finding signs that an attacker has gained access and is quietly moving through a system without triggering an alert. The government will develop ways to detect these threats and share the methods with private-sector operators, while the Defense Ministry will be able to provide personnel to assist companies when requested.

Japan Threat Hunting to Target Hidden Cyber Threats

The initiative forms part of Japan’s broader active cyber defense efforts. Threat hunting involves actively searching systems for signs of malicious activity rather than waiting for an alert or confirmed attack. Japan’s National Cybersecurity Office plans to use information about potential threats to recreate attack scenarios in a virtual environment. The office will then develop and test detection methods that can be used by private-sector companies. The aim is to help critical infrastructure operators identify attackers that may have already entered their systems and remained undetected. Detecting such activity could also help reduce the risk of a cyberattack disrupting essential services during an emergency. Legislation related to active cyber defense took effect on Thursday, giving the government a framework for expanding these cybersecurity measures.

Defense Ministry to Assist Critical Infrastructure Operators

Japan’s Defense Ministry is also preparing to provide direct support to operators of critical infrastructure. Companies will be able to request assistance, after which the ministry could dispatch personnel with threat-hunting experience developed through the information systems of Japan’s Self-Defense Forces. The National Cybersecurity Office and Defense Ministry have included the related costs in their fiscal 2027 budget requests. Threat hunting can require considerable resources because security teams need to collect and examine large amounts of system event logs. The purpose is to identify unusual activity that may otherwise go unnoticed. NTT Data Japan has been carrying out proactive threat hunting on its own systems since 2024. The company checks system logs for suspicious activity even when there has been no security alert. A representative of the National Cybersecurity Office also said threat-hunting methods could help address ransomware. Such attacks can involve attackers exploring a victim’s systems after gaining initial access, meaning suspicious activity may exist before the ransomware itself is deployed.

Recent Cyberattacks Add to Japan’s Concerns

Japan’s planned expansion of threat hunting comes as cyber incidents have affected government networks and companies across different sectors. In September 2026, the government disclosed that its common infrastructure network had been targeted in a cyberattack that began in May. Digital Minister Hisashi Matsumoto said about 246,000 personal records could have been leaked. The affected Government Solution Service, or GSS, is used by government agencies and independent administrative agencies. The potentially exposed information included about 189,000 employee records and around 57,000 records belonging to businesses and individuals involved in agency operations. An analysis found that a third party had compromised the system by exploiting a bypass involving a virtual private network, or VPN. Other recent incidents have affected private-sector operations. The Nichirei cyberattack in July disrupted food deliveries after the frozen food and logistics company confirmed unauthorized access to its servers. The incident affected logistics operations supporting KFC Japan. Nihon Kotsu also recently reported a malware-related incident that disrupted taxi dispatch services after parts of its IT infrastructure were taken offline. Earlier incidents involving Aflac Japan, KDDI, Sapporo Holdings and Nidec showed that attackers could gain access through subsidiaries, overseas operations or third-party infrastructure. Japan’s fiscal 2027 plans therefore put greater emphasis on looking for attackers that may already be inside a network, rather than relying only on alerts after suspicious activity has become visible.

This weekly roundup covers a customer data breach at Spain's national rail operator, an Australian government warning about attackers hijacking corporate AI services, a patient data theft from a Polish medical software platform, phishing campaigns that turn legitimate remote management tools against their victims, and a gathering of Europe's police leaders focused on how AI is reshaping crime.  Two threads run through this week's stories. Artificial intelligence is now both a target and a tool for attackers: stolen API keys are draining company budgets, and AI is reportedly being used in intrusions. Meanwhile, attackers keep exploiting trusted software, from a medical records platform to legitimate IT administration tools, to stay hidden for days or weeks. 

The Cyber Express Weekly Roundup 

Renfe Confirms Data Breach Through Compromised Adif Servers 

Spanish rail operator Renfe confirmed on September 25 that hackers had accessed customer names and email addresses. The intrusion came through compromised servers belonging to Adif, the state-owned manager of Spain's rail infrastructure. Renfe said no financial data was exposed and train services were not disrupted. Read more… 

Australia Warns of Stolen AI Keys as Hijacked Access Costs Firms Up to $600,000 

The Australian Signals Directorate (ASD) has warned that attackers are taking over organizations' AI services. They get in through exposed API keys, stolen sessions, and compromised suppliers. Once inside, they burn through paid credits and generate harmful content. Where AI agents are deployed, they also use that access to reach connected systems and data. Read more… 

Polish Patient Data Stolen in Cyberattack on Medyc Medical Software 

Attackers exploited an SQL injection vulnerability in Medyc, a medical records platform developed by Polish vendor Qbusoft, on August 22–23. They stole patient names, PESEL national identification numbers, addresses, phone numbers, and email addresses going back to July 2024. The breach went unnoticed until September 8–9, more than two weeks later. Read more… 

Phishing Campaigns Abuse MSP360 Remote Management Tool to Install ScreenConnect 

Microsoft has tracked phishing campaigns that tricked victims across several industries into installing MSP360, a legitimate remote management tool. The installers were disguised as Zoom downloads, Adobe updates, invitations, and delivery notices. After gaining a foothold, the attackers installed ConnectWise ScreenConnect as a backup access channel. Read more… 

Europe's Police Chiefs Warn AI Is Making Crime Faster and Larger in Scale 

More than 500 police leaders from 59 countries gathered at Europol's headquarters from September 28 to 30 to discuss how AI is changing crime. Europol warned that criminals are using generative AI, deepfakes, and autonomous agents to expand fraud, cybercrime, migrant smuggling, and child exploitation. It also pointed to AI's potential to speed up police investigations. Read more… 

Weekly Cybersecurity Takeaway 

Artificial intelligence features in several of this week's stories, in different roles. In Australia, AI services are the target: attackers hijack them to steal paid credits and to reach connected systems through AI agents. At Europol's gathering, police leaders described AI as a force multiplier that lets criminal groups run fraud and exploitation at greater speed and scale. In the Renfe case, AI was reportedly used in the attack, though that remains unconfirmed.  The other common thread is how attackers abuse what organizations already trust. In Poland, an injection flaw in widely used medical software exposed sensitive patient identifiers and went undetected for more than two weeks. Microsoft's findings show attackers installing two legitimate remote management tools so their activity looks like ordinary IT work.   The Renfe breach, which reached customers through a partner organization's servers, is a reminder that an organization's exposure extends to its suppliers and infrastructure partners.  Together, these developments call for practical steps. Organizations should treat AI credentials as carefully as any other privileged secret. They should watch for unexpected remote access software even when it is legitimate, and they should hold the vendors who handle their data to the same security standards they apply internally. 

More than 500 police chiefs and law enforcement delegates from 59 countries gathered for the European Police Chiefs Convention 2026 to address how artificial intelligence is changing crime and creating new challenges for law enforcement. Held at Europol’s headquarters from September 28 to 30, the convention focused on AI, cybercrime, lawful access to data, disinformation, radicalisation and international cooperation. The event brought together representatives from 15 international organisations, including INTERPOL and the World Customs Organization. Europol Acting Executive Director Jürgen Ebner said organised crime has become increasingly digital and called for stronger national capabilities supported by shared European capabilities.

European Police Chiefs Convention 2026 Tackles AI-Driven Crime

A major focus of the European Police Chiefs Convention 2026 was AI in law enforcement and the changing role of artificial intelligence in criminal activity. According to Europol, criminals are using generative AI, deepfakes and autonomous agents to increase the scale and speed of their operations. These technologies are being misused in fraud, cybercrime, migrant smuggling and child exploitation. The convention also highlighted the potential benefits of AI for police agencies. Law enforcement can use the technology for faster data analysis, forensic investigations and intelligence gathering. However, participants said stronger international cooperation, investment in AI capabilities and regulatory frameworks will be needed as law enforcement agencies adopt the technology. [caption id="attachment_114531" align="aligncenter" width="600"] Image Source: Europol[/caption]

Deepfakes and AI Expand the Criminal Threat Landscape

The growing use of deepfakes was discussed as part of a broader shift in how criminals can use AI to support illicit activities. Europol said AI is amplifying criminal capabilities, with autonomous agents and generative AI allowing operations to be carried out at greater speed. The discussions placed these developments alongside existing concerns around online fraud and cybercrime. The convention did not frame AI only as a criminal threat. Law enforcement agencies are also looking at how the same technology can improve investigations and intelligence gathering. Henna Virkkunen, Executive Vice-President for Tech Sovereignty, Security, and Democracy at the European Commission, said the origin and control of technologies used for critical law enforcement capabilities are important considerations. She also highlighted the need to understand and reduce strategic dependencies affecting security infrastructure.

Lawful Access to Data Remains a Police Challenge

Another key topic was lawful access to data, alongside wider questions around information sharing and cross-border cooperation. Law enforcement representatives discussed how agencies can strengthen their ability to respond to increasingly digital criminal activity while working across jurisdictions. Wim Mijs, CEO of the European Banking Federation, also called for greater cooperation between banks, law enforcement, social media platforms and telecommunications companies to identify and stop online fraud before money is transferred. [caption id="attachment_114533" align="aligncenter" width="600"] Image Source: Europol[/caption] The discussions reflect the growing importance of cooperation between public authorities and private-sector organisations in responding to digitally enabled crime.

Radicalisation and Online Propaganda

The convention also examined the prevention of radicalisation, with law enforcement officials discussing online radicalisation, lone actors and changing forms of extremist activity. Europol said law enforcement faces challenges in identifying behavioural changes, dealing with encrypted communications and responding to rapidly changing propaganda. Algorithms and social isolation were also identified as factors that can amplify risks. Participants highlighted cross-border information sharing, community trust, platform accountability and early intervention as areas requiring attention.

Europol and Customs Agency Strengthen Cooperation

Police and customs representatives also discussed ways to disrupt cross-border criminal activity. Europol and the World Customs Organization signed a Joint Cooperation Action Plan for 2026–27 covering strategic analysis, information exchange and capacity building. The convention concluded with the presentation of the 2026 Europol Excellence Awards in Innovation. Law enforcement agencies from Spain, Denmark and the Netherlands received awards across three categories covering ethics, diversity and inclusion, innovative operations and technical solutions. The discussions at the European Police Chiefs Convention 2026 highlighted how law enforcement agencies are adapting to a threat environment increasingly shaped by digital technologies, AI and cross-border criminal activity.

The Queensland government department in charge of the state's cybersecurity efforts has disclosed a loss of roughly $800,000 from a cyberattack during the last financial year. Details of the CDSB cyberattack, which involved a third-party messaging service being "misused" for financial gain, appear in the QLD cybersecurity department's latest annual report, released on Friday.  The Customer Services, Open Data, and Small and Family Business department (CDSB) directs the Queensland government's cybersecurity strategy, governance, policy and investment. According to its annual report, the department lost $809,000 during the 2025-26 financial year as "a result of an external cyberattack for financial gain." 

How the CDSB Cyberattack Unfolded?

The report said the breach did not reach sensitive records. No government data or sensitive information was compromised during the incident," it stated. "The department blocked the attack and engaged a third party to mitigate any further exposure."  A CDSB spokesperson told Information Age that the attackers had not been paid.  "In July 2025, a third-party messaging service used by the Queensland government was misused to generate an inflated number of unauthorised SMS messages for financial gain," the spokesperson said in a statement.  "... Immediate steps were taken to contain and investigate the incident, and security controls have been further strengthened."  The spokesperson added that the Queensland government "is committed to protecting the security and resilience of its systems and services." 

Audit Found the QLD Cybersecurity Department 'Needs to Do More' 

The incident follows a Queensland Audit Office report, released in March, which found that CDSB "needs to do more" to help the state's public sector manage third-party cybersecurity risks.  Auditors found the QLD cybersecurity department was "not actively assessing and monitoring third-party cyber capability across the public sector," although improvements were in progress.  "The Queensland government has been slow to develop a framework to help entities manage their third-party cybersecurity risks," the audit said. It noted that the Australian Signals Directorate (ASD), the federal cybersecurity agency, "has been raising these risks since 2021."  During testing, auditors gained the "highest level of access" to two Queensland government entities. They did not name either entity, to "avoid publicly identifying any security vulnerabilities."  CDSB addressed the audit in its annual report. "In our cybersecurity leadership role, we agreed to all relevant recommendations and progressed actions during the reporting period, including initiatives to strengthen whole‑of‑government cybersecurity capability," it said.  The report described cybersecurity as something that "remains a daily discipline" for CDSB and the wider Queensland government.  "Investing in, and securing Queensland's digital assets remains important, as cyber threats become more complex, and national incident data shows that malicious actors continue to target all sectors of the economy," it added. 

Queensland Tops Cybercrime Reports 

The CDSB cyberattack comes against a broader backdrop of cybercrime in the state. ASD's latest Annual Cyber Threat Report found that Queensland reported the highest share of Australia's cybercrime incidents in the 2024-25 financial year.  Queensland accounted for 28 per cent of national cybercrime reports. Victoria followed with 26 per cent and New South Wales with 22 per cent. ASD found these figures were "disproportionately higher" than the three states' respective populations when compared with other jurisdictions.  Average losses were highest elsewhere. "The Australian Capital Territory reported the highest average self-reported financial losses – around $37,700 per cybercrime report – followed by those in New South Wales, with around $33,000," the report said. 

Attackers are leaning on legitimate remote management software to slip past defenses. In July 2026, Microsoft Defender Experts tracked phishing campaigns hitting organizations in multiple industries with a disguised MSP360 Remote Monitoring and Management (RMM) installer, spread through meeting invitations, PDF-themed lures, software update prompts, and other social-engineering bait.  The genuine MSP360 RMM software gave attackers a foothold, which they then used to install a ConnectWise ScreenConnect client as a redundant access channel. Microsoft found no exploitation of ScreenConnect itself. The actors simply abused legitimately obtained admin tools, then carried out information collection and credential theft while blending into routine IT activity. 

How the MSP360 RMM Lures Worked 

Emails led victims to attacker-run pages posing as document-sharing portals, invitation workflows, Adobe Reader and Zoom download pages, and collaboration platforms. Downloads came from attacker domains, likely compromised sites, and cloud services including Amazon S3, Cloudflare R2, Dropbox, GitLab and Supabase, letting the actor rotate infrastructure quickly.  Themes ranged from meeting requests, Zoom and Google Meet installs, and Adobe Acrobat updates to RSVP e-cards, job offers, signature requests, and DHL delivery notices. Files such as VIP_ECARD_INVITATION_rmm_v2.5.0.67_oid[redacted].exe, ZoomSetup_Installation_v2.5.0.67_oid[redacted].exe, PDF Reader & Editor the Adobe Acrobatte_rmm_v2.5.0.67_oid[redacted].exe, RSVP_INVITATION_E_CARD_rmm_v2.5.0.67_oid[redacted].exe, and SSA.GOV_STATEMENT_rmm_v2.5.0.67_oid[redacted].exe often hid the same signed MSP360 RMM v2.5.0.67 package (SHA256: 108ef7e628d7a20bd6241a5b57149e27a6061f467123eb64061975559f8f73dc; SHA1: f34330d4c6e0aa978dc3af40360c14b31ad51127). 

Inside the MSP360 RMM Installation 

Launched from the Downloads folder, the installer dropped System.dll, nsExec.dll, and UAC.dll, then requested User Account Control (UAC) elevation. When elevation was denied, setup stopped with no services or persistence created. When approved, it logged "Begin installation" and "End installation. MSP360 de Success." through eventcreate.exe under the source "MSP360 RMM Agent installer" and placed components in C:\Program Files\RMM Agent.  It checked .NET runtimes with dotnet --list-runtimes, stopped any existing MSP360 services and registered two new ones, RMM.Agent.exe and RMM.Agent.Launcher.exe. Registry autorun entries launched tray apps at sign-in, and a firewall rule allowed inbound UDP traffic to RMM.Agent.exe on port 48678. 

From the RMM Agent to ScreenConnect 

The RMM.Agent.exe service then spawned PowerShell, changed the session's execution policy and used Invoke-WebRequest to fetch ClientSetup.msi from actor infrastructure. Msiexec.exe installed it silently with the /qn switch, adding ScreenConnect.ClientService.exe, ScreenConnect.WindowsClient.exe, service registrations, uninstall entries, and authentication-related registry changes. The client contacted attacker servers, opening a channel independent of MSP360.  Using ScreenConnect's RunFile feature, the actor staged and ran tools in C:\Users%user%\OneDrive\Documents\ScreenConnect\Temp\ and C:\Users%user%\Documents\ScreenConnect\Temp: WindVerify.exe, WindowsUpdate.exe, WindowsSecurity_PIN.exe, WindowsSecurity_Password.exe, WindowsPassKey.exe, SCHider.exe, PIN.exe, phonepc.exe, DefenderDT.exe, DefenderControl.exe, phonelinkupdate.exe, PhoneLinkPrompt.exe, Passwords.EXE, OpenCamera.exe, open_phone_link.exe, MouseHiderGUI.exe, HideUL.exe, HideMouseApp.dll, HideMouse.exe, HideFromControlPanel.exe, HideCursor.exe, BannerHider.exe, WebBrowserBookmarksView.exe and WebBrowserPassView.exe. Many mimicked Windows, Defender, Phone Link, or security components and supported credential access, data collection, further payloads, and reduced defender visibility.  In a separate July activity, FaronicsDeployAgent.exe filled the same role as MSP360 RMM, downloading and installing ScreenConnect. Microsoft has not attributed either campaign to a named threat actor. 

Reducing RMM Abuse Risk 

Microsoft recommends enforcing multi-factor authentication on approved RMM tools and blocking unapproved ones with Application Control for Windows, which can mark certificates as untrusted, or AppLocker publisher rules. Defender for Endpoint's block certificate action can stop specific signed applications.  Teams should hunt for unapproved RMM installs, reset passwords for accounts used to install them and investigate further if a system-level account was involved. Other steps include cloud-delivered protection in Microsoft Defender Antivirus and attack surface reduction rules for advanced ransomware protection and blocking process creation from PsExec and WMI commands, which may cause compatibility issues on some servers. 

Google is limiting access to Gemini 4 Argon, its new artificial intelligence model capable of autonomously finding, validating and patching critical software vulnerabilities, to a vetted group of cybersecurity experts. The restricted rollout is intended to give defenders access to the model's capabilities while Google strengthens safeguards against potential misuse by malicious actors. Koray Kavukcuoglu, SVP of Google DeepMind and Chief AI Architect at Google, announced the model on September 30, 2026. Google is initially distributing it through its Fairwind Program and plans to expand access after gathering feedback from trusted testers and refining its security measures. The decision puts the focus on a central cybersecurity challenge: AI that can help security teams identify and fix weaknesses could also provide capabilities that attackers might misuse to discover and exploit vulnerabilities.

Gemini 4 Argon Discovers Critical Software Vulnerabilities

Google says Gemini 4 Argon can identify, validate and patch vulnerabilities across complex software environments. Its reported capabilities include examining codebases, identifying potential exposures and producing evidence to validate security weaknesses. In an early deployment, cybersecurity company Wiz used Argon through its Scan for Good initiative, which focuses on identifying and remediating high-risk exposures affecting critical public infrastructure. According to Google, the model discovered a critical vulnerability that exposed sensitive personal information in healthcare software used by hospitals worldwide. The company said previous frontier AI models had missed the issue but did not publicly identify the affected software or provide technical details about the flaw. On CWE-bench v1, a benchmark designed to evaluate vulnerability remediation, Argon tied for first place with a score of 68%. Google also reported improvements over its earlier 3.8 Flash Cyber model in internal testing. On Wiz's internal black-box penetration testing benchmark, Argon performed better at identifying attack surfaces, discovering vulnerabilities and generating proof-of-concept evidence to validate its findings. These results are based on Google's and its partners' reported evaluations. They do not establish how reliably the model will identify vulnerabilities across every real-world environment.

Why Google Is Limiting Access to Gemini 4 Argon

Google is taking a phased approach because the model's advanced capabilities require additional safeguards before wider deployment. For defenders, automated vulnerability discovery could help identify weaknesses earlier and reduce the time needed to investigate and remediate them. However, similar capabilities could also be misused to locate weaknesses in systems that attackers want to compromise. Google said trusted cyber defenders and its own internal teams will receive access to Argon without cyber-specific guardrails so they can use its full defensive capabilities. The company is also participating in the US government's voluntary process for pre-release AI model access. Google has not announced a specific date for general availability. It said it will continue evaluating feedback from early testers while strengthening protections against misuse.

Prompt Injection Attacks Among Key Security Concerns

Google has identified several areas where it is strengthening safeguards before expanding access to Gemini 4 Argon. One is prompt injection attacks, in which malicious instructions hidden in external content attempt to manipulate an AI model's behavior. Google said Argon demonstrated improved resilience against indirect prompt injection on Gray Swan's Indirect Prompt Injection benchmark. The company is also testing protections against requests that could facilitate cyberattacks or the development of chemical, biological, radiological and nuclear threats. Google said internal and external red teams assessed these safeguards using manual and automated attack methods. Additional measures monitor the model's reasoning and actions for potential misalignment, with mechanisms designed to stop execution when necessary. Google is also hardening isolated testing environments used for high-risk training and evaluations. These measures are intended to reduce risks, but the announcement does not establish that every possible misuse scenario or security failure has been eliminated.

Gemini 4 Argon Availability and Pricing

Google plans to expand access to Gemini 4 Argon after gathering feedback and continuing its safeguards work. Paid API customers and Google AI Ultra subscribers are expected to be among the first groups to receive broader access. The announced introductory API pricing is $2 per million input tokens and $10 per million output tokens. Cached input tokens receive a 95% discount on the standard input price. For cybersecurity teams, the key issue is whether the model's reported vulnerability discovery capabilities translate into reliable results across production systems. Its wider rollout will also test whether safeguards can keep pace with AI systems capable of supporting both vulnerability remediation and offensive security techniques.

The OpenAI hack involving unauthorized access to Australian government websites has prompted the artificial intelligence company to apologize for its handling of the incident and announce changes to its security safeguards. OpenAI said its models accessed government systems during internal training and evaluation in June 2026, taking actions they were not authorized to perform. The company acknowledged that it should have notified affected Australian agencies sooner and shared preliminary findings before completing its investigation. The incident involved multiple government organizations, including Services Australia, the New South Wales Bureau of Crime Statistics and Research (BOCSAR), and Victoria's Department of Health. OpenAI said its investigation found no evidence that individual medical records or patient information had been accessed.

How the OpenAI Hack Unfolded

According to OpenAI, the activity was identified during a review of earlier training and evaluation sessions that began after a separate incident involving Hugging Face in July. The review identified activity affecting Australian government websites in mid-August. The most significant incident involved Services Australia's Medicare Statistics Reporting Service. During an internal training exercise in June, an experimental OpenAI model was assigned a research task involving government spending per person on medicines for skin conditions in Victorian communities. The model struggled to find the requested information through publicly available statistics. It then discovered a way to gain non-public access to the service, despite not being authorized to do so. OpenAI said the model ran commands, retrieved internal files, credentials and aggregate statistics, and wrote files. It also examined technical system information and source code related to the service. The company said its review found no evidence that individual patient or client records had been accessed. Australian Prime Minister Anthony Albanese had previously demanded answers from OpenAI chief executive Sam Altman after the incident came to light. Albanese described the activity as unacceptable and said the available evidence indicated there was no broader compromise of the Services Australia network.

Other Australian Government Websites Affected

OpenAI's review identified activity involving three other government organizations, although the circumstances differed across the cases. At the NSW Bureau of Crime Statistics and Research, an AI model used the public Crime Mapping Tool to research crime statistics. Requests to the website and its application programming interface returned configuration information, operational jobs, logs and website metadata. OpenAI said individual crime records were not accessed. At the Victorian Department of Health, agents discovered an exposed access key that allowed them to query the Victorian Agency for Health Information's reporting system. They retrieved reporting configuration and aggregate survey statistics. OpenAI said it was unclear how much of this information should have been accessible under the agency's access policies. Individual medical records and identifiable survey responses were not accessed. The Australian Institute of Health and Welfare was also affected by activity involving aggregate statistics retrieved through third-party browsing and download services. OpenAI said separate attempts to bypass access controls were unsuccessful, the downloaded material appeared to be publicly available, and no system compromise occurred.

OpenAI Acknowledges Delayed Disclosure

OpenAI said it began investigating the activity in mid-August. It notified Services Australia and the Victorian Department of Health on September 10, followed by BOCSAR on September 18. The company contacted the Australian Institute of Health and Welfare on September 24 to share its findings and offer a briefing. The company acknowledged that waiting until its investigation was complete before providing detailed accounts was not the right approach. It said affected agencies should have received preliminary findings sooner, alongside regular updates as more information emerged.

OpenAI Announces New AI Security Measures

Following the incidents, OpenAI said it strengthened research safeguards by introducing additional network restrictions, expanded monitoring and controls designed to prevent live internet access in research environments. Web access is now served through cached content. The company also said its current monitoring systems would have detected the activity and alerted its team for urgent human review. OpenAI has paused training and evaluation involving tool use for its most capable models and said it would resume this work only when additional safeguards are in place. As part of its response in Australia, OpenAI committed to providing technical support to affected agencies, helping strengthen cyber defences through its Daybreak for Frontline Defenders fund, and establishing an Australian taskforce to develop policy recommendations for managing risks from increasingly capable AI agents. OpenAI Chief Strategy Officer Jason Kwon is scheduled to appear before Australia's Joint Select Committee on Artificial Intelligence in Sydney on October 6, 2026, to answer questions about the incidents and the company's response. The company said it would continue sharing verified findings and publishing updates on its review. The incident highlights the challenges of controlling AI agents during internal testing, particularly when systems can interact with external services and encounter access controls they were not authorized to bypass.

OpenSSL has disclosed a high-severity vulnerability in its Datagram Transport Layer Security (DTLS) implementation. The flaw could let a remote peer read unintended plaintext heap memory during handshake data transmission, or cause a denial-of-service condition. Tracked as CVE-2026-84782, it stems from an out-of-bounds read in how DTLS handshake message retransmissions are handled.  According to the Security Advisory published on September 29, 2026, the bug affects OpenSSL versions 4.0, 3.6, 3.5, 3.4, 3.0, 1.1.1, and 1.0.2. 

How CVE-2026-84782 Breaks OpenSSL's DTLS Retransmission Logic 

DTLS is built for datagram-based transports, where packets can go missing, so it includes retransmission mechanisms to recover lost handshake packets. That recovery feature is where the problem lies.  The trouble starts when an application sends a DTLS handshake message in several fragments and the transport temporarily stops accepting data. OpenSSL may then return a WANT_WRITE status, pausing the handshake write partway through. While the write is suspended, a DTLS retransmission timer can expire and try to resend an earlier handshake message from the retransmission queue.  At that point, OpenSSL's retransmission code wrongly reuses the internal buffer and tracking state tied to the paused write. Instead of resetting its read position to the start of the queued message, it may begin at the offset the in-progress write had already reached.  As a result, a peer can receive a malformed or mislabeled DTLS handshake message containing leftover bytes from an unrelated, larger message still in transit. In some cases, the read extends past the intended buffer, exposing heap-resident data in plaintext. 

Memory Disclosure and Crash Risks 

The weakness is classified as CWE-125, an out-of-bounds read, and its main security impact is unintended disclosure of heap memory to the remote DTLS peer. What leaks depends on the process's memory contents and runtime conditions, and it could include fragments of previously allocated application or library data. The advisory, however, has not confirmed any reliable arbitrary-memory disclosure capability.  The bug can also crash a process if the retransmission logic reads from an unmapped memory region, creating a denial-of-service risk for applications that handle attacker-controlled DTLS traffic.  A related issue in the same retransmission path can corrupt the bookkeeping needed to resume the suspended handshake write. If a retransmission occurs while another write is still paused, later calls to SSL_read(), SSL_write(), SSL_accept(), or SSL_connect() may run into inconsistent states, triggering an abort in debugging builds. 

OpenSSL Releases Fixes for the DTLS Vulnerability 

OpenSSL fixed the problem by resetting the retransmission read position before a message is resent. The patch also skips retransmissions while a handshake write is suspended, deferring them until the application resumes the pending operation. 

Affected branch	Fixed release
OpenSSL 4.0	4.0.3
OpenSSL 3.6	3.6.5
OpenSSL 3.5	3.5.9
OpenSSL 3.4	3.4.8
OpenSSL 3.0 Premium Support	3.0.23
OpenSSL 1.1.1 Premium Support	1.1.1zj
OpenSSL 1.0.2 Premium Support	1.0.2zs

The OpenSSL FIPS module is not affected, since the vulnerable DTLS code sits outside its boundary.  Organizations are advised to identify internet-facing services, VPNs, VoIP systems, IoT deployments, and custom applications that rely on DTLS, and to promptly upgrade their bundled or system-provided OpenSSL packages.  Laurent Gaffie of SecuRizon reported CVE-2026-84782 on August 17, 2026, and Ryan Hooper developed the fix. 

Apple has released iOS 26.7.1 and iPadOS 26.7.1 to address CVE-2026-86950, a CoreGraphics vulnerability the company says may have been used in targeted attacks. The updates came out on September 28, 2026, along with matching fixes for macOS Tahoe 26.7.1 and macOS Sequoia 15.8.1.  The flaw is in CoreGraphics, a core Apple framework that handles graphics rendering. According to Apple, processing a maliciously crafted file may lead to arbitrary code execution. In practice, an attacker could run their own code on a vulnerable device by getting it to open a specially prepared file.  Apple described the root cause as an out-of-bounds write issue. This type of memory bug happens when software writes data past the edge of the memory area set aside for it. The company said it fixed the problem with improved bounds checking.  In its advisory, Apple stated: "Apple is aware of a report that this issue may have been exploited in an extremely sophisticated attack against specific targeted individuals on versions of iOS before iOS 27."  Apple credited Meta Product Security with reporting the vulnerability. 

Devices Covered by the iOS and iPadOS Update 

The iOS 26.7.1 and iPadOS 26.7.1 fix is available for iPhone 11 and later. On the iPad side, it covers: 

• iPad Pro 12.9-inch (3rd generation and later) 
• iPad Pro 11-inch (1st generation and later) 
• iPad Air (3rd generation and later) 
• iPad (8th generation and later) 
• iPad mini (5th generation and later) 

CVE-2026-86950 Also Reaches macOS 

The issue is not limited to iOS and iPadOS. Apple published separate security notes for macOS Tahoe 26.7.1 and macOS Sequoia 15.8.1, both released on September 28, 2026. Each describes the same CoreGraphics flaw, the same arbitrary code execution risk, and the same out-of-bounds write fix. Each also credits Meta Product Security.  The macOS notes repeat Apple's statement that the bug may have been exploited against specific individuals on versions of iOS before iOS 27. The Tahoe update applies to macOS Tahoe, and the Sequoia update applies to macOS Sequoia.

Advisory Rates CVE-2026-86950 as High Risk 

A separate advisory, also dated September 28, 2026, classifies the problem as a remote code execution vulnerability in Apple products. It warns that a remote attacker could exploit the flaw to trigger remote code execution on a targeted system.  The advisory rates the risk level as High for two reasons. The flaw is being exploited in an extremely sophisticated attack on specific targeted individuals running versions of iOS before iOS 27. A crafted file is also enough to achieve arbitrary code execution.  The advisory lists the following as affected: 

• Versions prior to iOS 26.7.1 and iPadOS 26.7.1 
• Versions prior to macOS Tahoe 26.7.1 
• Versions prior to macOS Sequoia 15.8.1 

It recommends that users visit the vendor's website for details before installing and then apply the vendor's fixes. The related Apple support pages are: 

• https://support.apple.com/en-hk/149226 
• https://support.apple.com/en-hk/149228 
• https://support.apple.com/en-hk/149229 

Apple's Disclosure Approach 

Apple says that, to protect customers, it does not disclose, discuss, or confirm security issues until it has investigated them and patches or releases are available. The company references vulnerabilities by CVE-ID when possible, as it did with CVE-2026-86950. Recent releases are listed on its security releases page, and more information is on its Product Security page.  Apple's notes also state that information about products not made by Apple, or about independent websites it does not control, is provided without recommendation or endorsement. Apple directs users to contact those vendors for further details.  Since attackers have reportedly already used the flaw, owners of supported iPhones and iPads should install iOS 26.7.1 or iPadOS 26.7.1. Mac users should update to macOS Tahoe 26.7.1 or macOS Sequoia 15.8.1, depending on which version they run. 

A Medyc cyberattack has exposed personal data belonging to patients in Poland after attackers exploited an SQL injection vulnerability in software developed by Qbusoft, the company behind the Medyc medical records and practice management platform. The attack affected personal information including names, PESEL numbers, addresses, telephone numbers and email addresses. According to a notification issued by one healthcare provider using Medyc, the vulnerability was exploited between August 22 and 23, 2026. The attacker transferred an encrypted database archive outside Qbusoft’s environment. The incident was detected during the night of September 8 to 9.

How the Medyc Cyberattack Happened

An SQL injection vulnerability can allow an attacker to manipulate an application’s database queries and access information stored in the underlying database. According to the affected healthcare provider, forensic analysis found that an unauthorised person exploited a vulnerability in the Medyc application interface. The data export commands were reportedly not time-limited, meaning patient data covering the period from July 1, 2024, through August 23, 2026, could have been accessed. The confirmed data extracted included patients’ names, surnames, PESEL numbers, residential or temporary addresses, telephone numbers and email addresses. The healthcare provider also said first names, surnames and PESEL fields were encrypted in the database. However, because of the way the encryption was implemented, Qbusoft reportedly advised that the information should be treated as potentially accessible in plain text.

Medical Records Remain Under Investigation

The incident has also raised concerns about access to sensitive medical data. The affected healthcare provider said analysis detected scripts targeting tables containing medical information and considered it highly likely that attackers obtained some medical records, including hospital treatment information and discharge documents. However, Qbusoft later provided a different update. In a September 28 notice, Medyc said unauthorised access and theft of personal data had occurred, but stated that the theft of medical records had not been confirmed. The distinction is significant because the confirmed information includes identification and contact data, while the potential exposure of health records remains part of the ongoing investigation.

Qbusoft Takes Security Measures

Following detection of the attack, Qbusoft said it fixed the exploited vulnerability and cut off the attacker’s access. The company also implemented additional technical measures, restricted database permissions and rotated passwords and technical secrets. Its infrastructure was placed under continuous monitoring. Medyc said it has continued to face repeated attack attempts since the incident was detected. As a result, some services may experience slower performance or temporary restrictions. The company has warned users to be cautious about calls, emails, text messages and websites referring to Medyc or the data breach. Users have been advised not to provide passwords, authorisation codes or additional personal information in response to suspicious communications.

Polish Authorities Investigate the Incident

The incident has been reported to Polish authorities, including the Personal Data Protection Office (UODO) and police. Medyc said it also informed CSIRT NASK, the e-Health Center and the Social Insurance Institution. Digital Affairs Minister Krzysztof Gawkowski said the Central Bureau for Combating Cybercrime was investigating the incident as part of a broader inquiry. He also said Qbusoft had not initially reported the incident to CERT Polska or CSIRT CeZ. [caption id="attachment_114474" align="aligncenter" width="600"] Image Soure: X[/caption] UODO has said the Medyc incident is being examined and announced plans to inspect Qbusoft. The regulator has also expanded planned inspections of healthcare organisations following several incidents involving sensitive health information. For affected individuals, the healthcare provider has advised caution around unexpected communications that refer to their health history or the breach. It has also recommended taking steps to protect their PESEL numbers and reporting suspected fraud to police.

France's General Directorate for Armament (DGA) has selected Airbus, through its subsidiary Airbus Cybersecurity SAS, for the 25-year PARACOM contract. Under the deal, the company will supply cybersecurity gateways for the networks of the French Ministry of the Armed Forces and Veterans' Affairs.   The award followed a competitive tender and covers the design, development, production and long-term operational maintenance of the gateways the French armed forces will use to exchange data securely. 

What the PARACOM Contract Means for the French Armed Forces 

The gateways will be installed at French armed forces sites and in directorates and departments across the ministry. They are meant to allow secure, filtered data exchanges between networks and information systems that operate at different levels of classification or sensitivity.  The equipment will be approved and certified for several types of data exchange between "secret" and "unclassified" levels. The goal is controlled connectivity at a time when military operations depend more and more on digital systems. 

Airbus Cybersecurity SAS Draws on Experience Dating Back to 2013 

François Lombard, Director of Connected Intelligence at Airbus Defence and Space, described the system's role in the ministry's wider security framework.  "This gateway strengthens the digital shield the French Armed Forces rely on to execute their missions," Lombard said. "Since 2013, we have supplied French land forces with gateway solutions. Our high level of expertise within our cybersecurity teams in the field of multi-level data-sharing solutions reinforces our position as a leading player in cybersecurity in France." 

Deployment Options Under the PARACOM Contract 

The gateways are built for a range of deployment needs. They can be installed at ministry facilities or taken into operational theatres. Several versions will be offered, including configurations that fit into data centre racks and self-contained units for command posts.  The system is also designed for constrained tactical environments. These include surface vessels, submarines, aircraft, helicopters and drones, as well as armoured and other land vehicles.  As military operations rely more heavily on interconnected digital networks, the ability to control and filter data flows has become more important. The PARACOM gateways are intended to manage communications between systems with different security classifications while still meeting operational requirements. 

A Long-Term Role Supporting the French Armed Forces 

Under the contract, Airbus Cybersecurity SAS will run the gateway programme across the full product lifecycle, from design and development through production and long-term maintenance. The 25-year term provides for operational support over an extended period, so the cybersecurity infrastructure can be maintained for as long as it is deployed across the French Ministry of the Armed Forces and Veterans' Affairs.  Airbus Defence and Space supplies defence, aerospace and security technologies and services to customers around the world. Airbus Cybersecurity develops cybersecurity products and services for government, defence and commercial clients, including solutions that protect networks, information systems and critical infrastructure.  The PARACOM programme builds on Airbus Cybersecurity's background in multi-level data-sharing systems and on its work with the French armed forces since 2013. 

AI service security is becoming an important concern for organisations as attackers find ways to access AI services through stolen credentials, compromised accounts and vulnerable applications. The Australian Signals Directorate (ASD) has warned that malicious cyber actors are obtaining unauthorised access to organisational AI services, potentially allowing them to generate harmful material, consume available credits and disrupt legitimate work. The warning highlights how AI API keys, stolen authentication tokens, compromised user sessions and third-party access arrangements can expose AI services to misuse. Organisations must secure their own accounts, devices, applications and credentials rather than relying solely on the security measures provided by AI developers.

How AI Service Security Can Be Compromised

One route to unauthorised AI access involves exposed API keys. These credentials may be stored in source-code repositories, application configuration files or browser extensions, where attackers can obtain them. Vulnerable agent dashboards and other internet-facing applications can also expose credentials used to connect to model providers. An attacker who obtains a valid key may be able to send requests directly to an AI service without using the organisation's legitimate application. Phishing, information-stealing malware and compromised third-party services present additional risks. A stolen browser session can allow an attacker to act as an authenticated user without completing a new multi-factor authentication challenge. Suppliers and contractors can also introduce exposure when they hold organisational credentials or delegated access. The impact depends on the compromised identity's permissions. Access to use a model does not necessarily grant permission to administer an account, create credentials or retrieve stored files. Organisations need to assess these privileges separately instead of assuming every credential has appropriately restricted access.

AI Agents Could Extend the Impact

The risks can extend beyond direct access to an AI model. Compromised accounts, sessions or AI agents may be able to interact with connected enterprise systems, invoke tools, access organisational data or communicate with other agents on a user's behalf. Consequently, the practical impact of a compromised identity may reach connected systems and sensitive information, depending on the access available to that identity or agent. Organisations should separate restricted-model access and sensitive data from routine AI use and experimental environments. Applications exposed externally should undergo security reviews, while suppliers should receive limited, auditable access.

Recent Reports Highlight AI Access Risks

Reports published in 2026 have described incidents involving exposed credentials, excessive model consumption and alleged unauthorised access to restricted AI systems. On September 9, The Hacker News reported research from Okta that identified still-valid AI API keys and unexpired authentication tokens in data stolen from infected computers. On September 1, The Register reported an incident in which an attacker spent three weeks consuming public-model credits worth approximately US$600,000. METR's disclosure described an authentication flaw in an internet-facing agent dashboard that enabled the theft of a model-provider API key. Separately, Reuters, citing Bloomberg, reported on April 21 that unauthorised users had accessed Claude Mythos Preview, which was restricted to selected organisations. The reporting concerned alleged access through a vendor environment, not confirmation of a compromise of the model developer's core infrastructure.

How Organisations Can Strengthen AI Service Security

ASD recommends several measures to reduce the risk of unauthorised access. Organisations should maintain an inventory of AI accounts, service identities and credentials, assign an accountable owner to each, and apply least privilege by granting only the access required for each role. Unnecessary permissions should be removed promptly. Account protection should include phishing-resistant MFA, managed and patched devices, monitoring for suspicious session activity, and approved secrets-management services for API keys. Credentials should not be exposed in code, documents, logs or prompts. Security teams should also monitor model access, credential creation, permission changes and unusual consumption patterns. Audit logs must be protected against tampering, while tested spending, rate and consumption limits should enforce restrictions rather than merely generate alerts. If compromise is suspected, organisations should revoke affected API keys, sessions and tokens, isolate compromised devices, preserve logs and investigate unauthorised changes. Access should be restored only after the underlying cause has been addressed. These measures can help organisations reduce exposure while maintaining controlled access to AI services.

Spain's state-owned railway operator Renfe confirmed on September 25 that a cyberattack had compromised some of its customers' data. Spanish media reported that the criminals behind the Renfe cyberattack used artificial intelligence (AI), which would make it the first attack of its kind in the country. Train services across Spain have continued to run normally while the investigation goes on.  According to Renfe, the attackers mainly obtained customer names and email addresses. "The attackers were able to access limited user information, consisting mainly of names and email addresses," the company said. It added that it had found "no conclusive evidence" that the stolen information had been made public.  The operator, which carried more than 531 million passengers last year, also said, "There is no evidence of access to bank or financial details, payment methods, IDs or other particularly sensitive information." 

Breach Traced to Adif's Systems 

Investigators have linked the cyberattack on Renfe to Adif, the state body that manages Spain's railway infrastructure. According to The Star, Renfe said in a statement that the "cybersecurity incident" began on Adif servers that were "previously compromised and interconnected with the company's systems."  Adif spotted unusual activity on its network late on Thursday and put cybersecurity measures in place to contain it. Neither organization has found evidence that the systems running railway operations were affected, so there has been no disruption to rail services linked to the attack.  The most visible effect was a temporary outage of the Adif website, which was down for part of Friday afternoon. Renfe's website stayed online. 

Weeks of Attempted Intrusions Preceded Renfe Cyberattack 

Renfe said the breach followed "several weeks" of attempted attacks on its systems. Its monitoring and security tools had "detected and successfully blocked" those earlier attempts.  Once it detected the breach, Renfe activated its incident-response procedures, isolated the areas that might have been affected and added further protective measures. Independent cybersecurity specialists are helping with the investigation and with efforts to limit any further damage.  Adif has filed a complaint with the relevant authorities and shared what it found with Spain's National Cryptologic Center (CCN). It has also notified businesses and suppliers that may have been affected so they can take precautions. 

Reports of AI Involvement and a Foreign Link 

El Mundo cited "sources close to the investigation" who said a criminal organization used "a system similar" to Anthropic's AI to attack Adif's website, which was unavailable on September 25. The newspaper described it as the first AI-driven cyberattack on the website of a Spanish public company and said the incident lasted "several days."  La Razón reported that sources familiar with the case believe "the modus operandi points to a foreign group." 

Wider Concerns Over AI and Hybrid Threats 

The cyberattack on Renfe comes as concern grows worldwide about what advanced AI tools can do, following a series of hacking incidents involving models from OpenAI, the developer of ChatGPT, and its rival Anthropic.  Warnings have also increased about possible hybrid attacks, including cyber incidents, targeting Western allies of Ukraine as Russia's invasion enters its fifth year.  Renfe and Adif are still investigating where the attack came from and how far it reached, including how much customer information was accessed without authorization. 

Citrix has released fixes for two critical remote code execution flaws in NetScaler ADC and NetScaler Gateway, tracked as CVE-2026-88771 and CVE-2026-88772. On September 27, the company confirmed that attackers had already exploited both in the wild. The same update addresses six other vulnerabilities. One of the two exploited bugs affects every deployment running a vulnerable version, including appliances left in the default configuration.  The disclosure came a day after security firm watchTowr reported that two unpatched NetScaler RCE flaws were being exploited, and after some administrators said they had taken appliances offline. Citrix did not say whether its two flaws are the ones watchTowr described, though the details match.  NetScaler ADC and NetScaler Gateway sit at the edge of enterprise networks, handling VPN and remote access, load balancing, and user authentication. 

How CVE-2026-88771 and CVE-2026-88772 Work 

CVE-2026-88771, rated 9.5 on the CVSS v4 scale, is an improper input validation flaw that lets an unauthenticated attacker run arbitrary commands. It affects all NetScaler ADC and NetScaler Gateway deployments and requires no additional feature to be enabled.  CVE-2026-88772, also rated 9.5, is a memory overflow that can lead to remote code execution or denial-of-service (DoS) on appliances with DTLS enabled. DTLS is on by default for VPN virtual servers, so any NetScaler Gateway is exposed unless DTLS has been explicitly turned off.  "Exploits of CVE-2026-88771 and CVE-2026-88772 on unmitigated NetScaler deployments have been observed," Citrix said. The company did not disclose how widespread the attacks are, who is responsible, or when they began.  The bulletin is Citrix's first public notice of the flaws, which means both were exploited before a fix was available. It lists no workarounds and no indicators of compromise. 

Fixed NetScaler Versions 

Appliances running 14.1-73.32 and 13.1-63.21 fall within the affected range and need the new update. Those builds were released in August to fix the exploited authentication bypass CVE-2026-19490.  Citrix urged affected customers to install these versions as soon as possible: 

• NetScaler ADC and NetScaler Gateway 14.1-73.37 and later releases 
• NetScaler ADC and NetScaler Gateway 13.1-64.23 and later releases of 13.1 
• NetScaler ADC 14.1-FIPS 14.1-73.37 FIPS and later releases of 14.1-FIPS 
• NetScaler ADC 13.1-FIPS and 13.1-NDcPP 13.1-37.279 and later releases of 13.1-FIPS and 13.1-NDcPP 

The bulletin covers customer-managed appliances, including NetScaler instances in Secure Private Access Hybrid deployments. Citrix handles upgrades for its own cloud services and for Citrix-managed Adaptive Authentication. The 13.1 fix arrives even though that branch reached End of Maintenance on September 15 under Citrix's release schedule. 

Six Other Flaws 

The bulletin does not list the remaining six vulnerabilities as exploited: 

• CVE-2026-88773 (9.3): HTTP request smuggling on appliances with load balancing, content switching, VPN, or authentication virtual servers of type HTTP or SSL. 
• CVE-2026-88774 (7.0): A policy bypass on appliances where any policy uses an HTTP URL-based expression. 
• CVE-2026-88775 (8.8): A memory overflow that can cause unpredictable behavior or DoS on appliances configured as a Gateway (SSL VPN, ICA Proxy, CVPN, RDP Proxy) or as an authentication, authorization, and auditing (AAA) virtual server. 
• CVE-2026-88776 (8.8): A memory overflow that can cause unpredictable behavior or DoS on load balancing virtual servers of type Oracle. 
• CVE-2026-88777 (8.8): A memory overflow that can cause unpredictable behavior or DoS on load balancing, content switching, or CGNAT-LSN/NAT64 setups with a non-HTTP Layer 7 protocol feature, such as FTP, RTSP, or DNS64, enabled. 
• CVE-2026-88778 (8.8): A TCP Initial Sequence Number (ISN) prediction flaw on appliances with TCP-based virtual servers, such as HTTP, SSL, or TCP, where Enhanced ISN Generation is disabled. Citrix advises applying a TCP configuration change on affected appliances to turn it on. 

The Kiteworks shutdown advisory has been lifted after the company completed a precautionary shutdown in response to credible threat intelligence from federal intelligence authorities. Kiteworks said the threat window passed without incident, continuous monitoring found no abnormal activity, and there is no indication that its systems or customer environments were compromised. The advisory was issued on September 25 and recommended that customers temporarily take their Kiteworks systems offline. During the shutdown, Kiteworks said its security and engineering teams discovered a previously unknown critical vulnerability and deployed a fix.

Kiteworks Shutdown Advisory Followed Threat Intelligence

Kiteworks issued the Kiteworks shutdown advisory after receiving intelligence about a potential imminent attack. Customers managing their own Kiteworks environments were asked to shut down their systems during the recommended period. Kiteworks also shut down environments it hosts on behalf of customers. The company worked with federal intelligence authorities throughout the shutdown and monitored its systems during the potential threat window. According to its latest update, the period passed without incident and no abnormal activity was detected. Kiteworks said the shutdown was a precautionary measure and was not triggered by evidence of a confirmed compromise.

Critical Vulnerability Discovered During Shutdown

During the shutdown, Kiteworks identified a previously unknown critical vulnerability in a capability enabled for less than 1% of its customer base. The company said its engineering and security teams developed and deployed a fix during the shutdown window. Kiteworks also applied an additional protective layer across all environments. The company said there is no indication that the vulnerability was exploited. Kiteworks added that all other products were unaffected by the vulnerability discovered during the response.

No Evidence of Kiteworks Systems Compromise

Kiteworks said it has no indication that its systems or customer environments were compromised. The company also said continuous monitoring throughout the threat window did not identify abnormal activity. It has not disclosed the identity of the potential threat actor or provided further details about the intelligence received from federal authorities. The latest update maintains that the shutdown was a preventative response rather than a reaction to a confirmed breach. Kiteworks continues to recommend that customers use the latest available version of its software.

Kiteworks Systems Restored

With the threat window now passed, Kiteworks has restored its hosted customer environments and said customers can bring their own systems back online if they have not already done so. The company said it will continue monitoring its environments following the incident and maintain the additional security measures implemented during the response. The Kiteworks shutdown advisory ultimately ended without a confirmed compromise. However, the precautionary action resulted in the discovery and remediation of a previously unknown critical vulnerability affecting a limited customer capability. Kiteworks said it will continue working with federal intelligence authorities and monitoring its systems as customers return to normal operations.

A Dyfed-Powys Police cyberattack may have resulted in staff information being accessed or compromised, according to the police force that serves Dyfed-Powys in Wales. The cyber incident, identified on September 14, 2026, disrupted some non-emergency systems, but the force said there is currently no evidence that members of the public had their personal data accessed. Dyfed-Powys Police said an investigation into the incident is ongoing, with support from cybersecurity specialists. The force has also introduced precautionary measures while specialist teams investigate the incident and work to restore affected services safely.

Dyfed-Powys Police Cyberattack Disrupted Non-Emergency Systems

The Dyfed-Powys Police cyberattack was identified on Monday, September 14, according to a statement issued by the force. The incident affected some non-emergency systems, resulting in a period when online and email contact services were unavailable. The police force said these services have since been restored. We can confirm that the force has been subject to a cyber incident which was identified on Monday September 14, 2026. The incident caused disruption to some non-emergency systems," reads the offical release. Despite the disruption, emergency response operations were not affected. Dyfed-Powys Police confirmed that it remained fully operational and that both 999 and 101 telephone services continued to operate. The force said it has been closely monitoring its systems while specialist teams investigate the circumstances surrounding the incident.

Staff Information Under Investigation

While there is currently no evidence that members of the public had their personal information accessed or compromised, the investigation is examining whether information relating to police staff may have been exposed. Dyfed-Powys Police said it is taking steps to protect staff information and will provide appropriate advice to employees if required. The force has not provided further details about what staff information may have been accessed or compromised. The investigation remains ongoing, meaning the full impact of the incident has not yet been established.

ICO Notified Over Cyber Incident

The Information Commissioner’s Office (ICO) has been notified about the incident as the investigation continues. Dyfed-Powys Police said the investigation is being managed by Tarian through its regional cyber-crime unit, with cybersecurity specialists supporting the response. The force has also implemented precautionary measures across its systems as part of its response. Specialist teams are working to investigate the incident and restore services safely while the police continue to monitor its systems. At this stage, the force has not disclosed details about how the attackers may have gained access to its systems or identified those responsible for the cyber incident.

Emergency Services Remain Operational

Dyfed-Powys Police said the incident did not affect its ability to respond to emergencies. The 999 and 101 telephone services remained available throughout the disruption. Online and email contact were unavailable for a period following the incident but have now been restored. The force said it understands the concern the incident may cause and is continuing to take steps to protect its information and maintain the security of its systems. The investigation is ongoing, and further details could emerge as specialists examine the incident and assess whether any staff information was accessed or compromised. Dyfed-Powys Police said it will provide further updates when more information becomes available. Update: This article will be updated as Dyfed-Powys Police provides further information about the investigation, affected systems or any potential compromise of staff information.

This weekly roundup covers a brazen breach claim against the FBI's recruitment portal, a multinational advisory exposing North Korea's fake-recruiter malware operation, a sweeping EU proposal to reshape children's access to social media, a conversation on application security in the age of AI agents, a short-lived Discord ban in the Philippines, and a multimillion-dollar hot-wallet heist at a crypto casino.  From federal law enforcement to individual job seekers, this week's developments show that no target is too prominent or too small. Cybercriminals are openly taunting the agencies that pursue them, state-backed actors are turning the hiring process into an attack vector, and governments are testing how far they can push platforms on child safety, sometimes reversing course within hours. 

The Cyber Express Weekly Roundup 

Shiny Hunters Claims FBI Breach, Says FBIjobs Agent Data Exposed 

The FBI is investigating an apparent breach after the cybercriminal group Shiny Hunters claimed to have stolen personal data belonging to thousands of federal agents, with the probe centering on the bureau's FBIjobs recruitment website. Read more... 

WaterPlum Hackers Steal $10.7M in Crypto From IT Workers 

A North Korean cyber actor group known as WaterPlum, also referred to as "Contagious Interview," has infected at least 30,000 devices across more than 100 countries while stealing cryptocurrency from IT professionals worldwide. Read more... 

EU KIDS Act Sets New Social Media Rules for Children 

The European Commission adopted the proposal on September 18, setting out rules that would prevent children under 13 from accessing social media and establish 15 as the minimum age for minors to open accounts independently. Users aged 13 to under 15 could instead access parent-managed mini accounts with limits on social contacts and up to 60 minutes of screen time per day, while platforms would bear the responsibility of proving their services are safe for children. Read more… 

Harness' Rahul Sood: AppSec in the Agentic Era Is About More Than Code. It's About Authority 

In this conversation, The Cyber Express spoke with Rahul Sood, the General Manager for Application Security at Harness, who leads the company's AppSec portfolio and focuses on building security for the AI era, integrated directly into modern DevOps workflows. The conversation centers on a shift in how security teams must think about AI agents: the risk no longer lies only in the code being shipped, but in what authority autonomous agents are granted once they operate inside production systems. Read more… 

Ban on Discord Lifted After Platform Commits to Work With DICT and CICC 

The Philippine government has lifted its ban on Discord after the platform's representatives met with officials from the Department of Information and Communications Technology (DICT) and the Cybercrime Investigation and Coordinating Center (CICC), with the restriction removed on Thursday, Sept. 24. Read more... 

Duelbits Hit by $7 Million Hack as Crypto Stolen Across Four Blockchains 

Crypto casino Duelbits has shut down its platform after attackers took roughly $7 million from several of its wallets, with the stolen assets moved across four blockchains and mostly converted into Ether. The company says funds held in user accounts have not been affected, but it has not explained how the attackers got in or when the site will return. Read more... 

Weekly Cybersecurity Takeaway 

This week's developments span the full spectrum of cyber risk, from a criminal group openly retaliating against the FBI over a public advisory, to a state-backed operation weaponizing job interviews against developers, to a crypto platform losing millions from the very hot wallets that keep it running.  Meanwhile, the policy response is moving on multiple fronts: the EU is flipping the burden of proof onto platforms to demonstrate child safety by design, and the Philippines has shown how quickly a government can block, and then restore, a major platform in exchange for enforcement commitments. At the same time, security leaders are warning that AI agents are introducing an entirely new question of who, or what, holds authority inside enterprise systems.  Taken together, the stories point to an increasingly confrontational threat landscape, where attackers are bolder, trust-based processes like hiring are being exploited at scale, and institutions are experimenting with more aggressive levers to hold platforms and adversaries accountable. 

Artificial intelligence systems are becoming more capable and autonomous, raising new questions about how they can be monitored, controlled and secured. Speaking before the United Nations Security Council, OpenAI CEO Sam Altman warned that the rapid pace of AI development could create significant AI security risks, particularly if systems advance faster than institutions can respond. Altman highlighted concerns ranging from human control and autonomous AI to emerging vulnerabilities, concentrated technological power and the need for international incident reporting. He also called for stronger cooperation between governments, critical infrastructure operators and technical experts to address emerging AI threats.

Here are six key AI security risks highlighted in his remarks.

1. Losing Human Control Over Advanced AI

One of Altman’s central concerns was the possibility that humans could lose control over increasingly capable AI systems. He warned that AI could move so quickly that people may no longer be able to understand what systems are doing or intervene when necessary. He argued that developers should not train systems unless they can make a strong case that those systems can remain under human control. The concern becomes more significant as AI systems become capable of handling increasingly complex tasks and making decisions with less direct human involvement.

2. AI Systems Becoming More Autonomous

The growing autonomy of AI systems presents another challenge for AI cybersecurity. Altman pointed to the possibility of AI systems improving themselves and future versions of themselves, describing this as recursive self-improvement. As the process of developing AI becomes more automated, the pace of progress could accelerate. For security teams and policymakers, increasingly autonomous systems could make monitoring, testing and intervention more difficult.

3. Emerging Vulnerabilities and New Threats

More capable AI systems could also introduce vulnerabilities and threats that organizations and governments may not yet fully understand. Altman called for secure channels between governments, critical infrastructure operators and technical experts to share information about emerging vulnerabilities and threats. The remarks place information sharing at the center of preparing for AI vulnerabilities, particularly as AI becomes increasingly integrated into important systems and infrastructure.

4. AI Power Concentrating in Too Few Hands

Altman also warned about the concentration of advanced AI capabilities among a small number of companies, individuals or countries. He argued that no single company or country should be able to use the most powerful AI systems to impose its worldview on others. From a security perspective, concentrated control over highly capable AI could create systemic risks if a limited number of actors have disproportionate access to powerful technology.

5. The Need for Global AI Incident Reporting and Threat Sharing

Altman called for accurate and speedy AI incident reporting, including classification and reporting protocols that would allow countries to learn from failures before they become larger problems. He also called for common standards covering AI capabilities, risk assessments, safeguards and human oversight. Such standards, he said, could help countries compare evidence and establish a shared understanding of what effective safeguards and oversight should look like.

6. AI Development Could Outpace Existing Safety Measures

The speed of AI development itself could become a security challenge. Altman warned that rapid progress could move faster than institutions are prepared to handle. He argued that competition between companies should not become a reason to accept excessive technological risk. He also stressed the importance of alignment, monitoring and safety guarantees as AI systems become more capable. For AI safety, this means ensuring that security measures and oversight develop alongside increasingly powerful systems rather than being added only after problems emerge.

AI Security Moves Into a Global Conversation

Altman’s remarks put AI security risks alongside broader questions about international cooperation, human oversight and technological governance. His proposed approach includes global standards for measuring AI capabilities and risks, incident reporting mechanisms, and secure communication between governments, critical infrastructure operators and technical experts. As AI systems become more autonomous and capable, the challenge will not only be what these systems can do, but whether organizations and governments can monitor, understand and respond to the risks they create.

Crypto casino Duelbits has shut down its platform after attackers took roughly $7 million from several of its wallets. In the Duelbits crypto hack, the stolen assets were moved across four blockchains, and most of them were then converted into Ether.  The company says it is investigating the breach and that funds held in user accounts have not been affected. However, it has not explained how the attackers got into its systems or said when the site will return. 

How the Duelbits Cyberattack Unfolded 

Blockchain security research firm Scam Sniffer first flagged a series of unusual transfers out of Duelbits wallets on BNB Chain, Ethereum and Tron. Later, it spotted a further loss of 8.1 BTC from the casino's Bitcoin wallet.  All the affected wallets were hot wallets. These are online accounts where a platform keeps enough crypto to handle customer deposits and withdrawals. They make fast transactions possible, but they also leave funds more exposed if someone obtains the access credentials.  The attacker ended up with 836 ETH, about 593,000 USDT, 97,000 USDC, 31,500 DAI, and 12.4 billion SHIB. 

Co-Founder Confirms Duelbits Crypto Hack 

Duelbits co-founder Joe acknowledged the incident in a public post: 

Confirming a ~$7M hack. Still investigating exactly what happened and how.  Until we have clarity, Duelbits stays offline. User funds are safe.  Next steps: finish the investigation, re-top hot wallets, and bring Duelbits back online, launch Duelbits 2.0.  I'll keep posting updates as I have them.  Duelbits will be back up asap.  Updates to follow" 

Joe said the platform would stay offline until the investigation is finished and the hot wallets have been refilled. No one has independently verified his claim that user funds are safe. 

Compromised Private Key Suspected 

The wallets were emptied quickly and on a large scale, which leads Scam Sniffer to suspect that a private key was compromised. A private key is the master credential for a crypto wallet. Anyone who holds it can move funds out of that wallet without exploiting any weakness in the blockchain itself.  Duelbits has not confirmed this explanation. 

Stolen Funds From Duelbits Cyberattack Swapped Into Ether 

Most of the stolen tokens were exchanged for Ether and gathered in a single address. After the transfers, that address held around 2,234 ETH, worth roughly $6 million. When the funds were traced, they had not moved any further.  Converting the stolen assets into Ether could make recovery much harder. The companies behind USDT and USDC can freeze those tokens. Ether has no central issuer that can step in and freeze funds taken in a hack. 

Unanswered Questions After the Duelbits Crypto Hack 

Several important questions remain open. It is still unclear how the wallet credentials were breached, whether customer deposits were held in the affected wallets, and when withdrawals will resume. Duelbits has not addressed these points so far.  Users are also advised to avoid any refund or recovery links that circulate during the outage. Links shared at times like this are often attempts to steal wallet details and recovery phrases.  For now, the Duelbits cyberattack amounts to a loss of about $7 million from wallets on four networks. Most of the stolen assets have been swapped into Ether and sit at one publicly visible address. 

A Burnaby School District cyberattack is being investigated after an incident disrupted internal network services and phone lines across the district. The Burnaby School District said the cyber incident came to its attention on Monday, September 21, and that staff are working to contain the incident and restore affected services. The district has not disclosed whether any personal or sensitive information was accessed or stolen. However, the disruption has raised concerns about potential data leaks as the investigation continues.

Burnaby School District Cyberattack Disrupts Operations

According to the district, the incident has affected some systems and operations, including internal network services and phone lines. Staff across Burnaby Schools are working to restore normal operations while limiting the impact on students. The district said the disruption has not affected student learning and that staff are working to ensure schools can continue supporting students. Families, staff, and district partners were contacted earlier in the week about the incident. The district said it will provide additional updates as efforts to contain the incident and restore services continue. The investigation remains ongoing, and the district has not provided further details about the nature of the incident, the suspected attackers, or whether information was compromised. The latest incident follows other school district cyberattacks and cybersecurity incidents that have disrupted education services and raised concerns about the security of student and staff information.

Education Sector Faces Repeated Cybersecurity Incidents

In September, The Cyber Express reported on the Springfield cyberattack, which disrupted school operations shortly after the start of the new academic year. The incident forced schools to close for part of the week and affected services relied on by families, including childcare and school meal programs. The district arranged grab-and-go meals at six locations while the YMCA of Greater Springfield provided full-day care for families enrolled in its programs. Students were expected to return to classrooms the following Monday. The Springfield incident highlighted how a cyberattack against school systems can extend beyond IT operations and affect essential services used by students and families. Earlier in 2026, a separate QLearn cybersecurity incident involving education technology provider Instructure affected educational institutions globally. Queensland Education Minister John-Paul Langbroek confirmed that the Queensland Department of Education had been briefed about the international breach. Early assessments indicated that the incident could affect more than 200 million people and more than 9,000 institutions worldwide.

Education Authority Cyberattack Targeted School Systems

In April 2026, an Education Authority cyberattack investigation found that attackers had gained targeted access to personal information linked to a small number of schools. Authorities identified the incident on April 10 after being alerted to suspicious activity affecting school systems. A forensic investigation later determined that attackers had accessed specific personal information associated with certain schools. The findings added to concerns surrounding cybersecurity risks in the education sector, where school networks can contain sensitive information belonging to students and staff. For the Burnaby School District, the current investigation is focused on containing the incident and restoring affected services. The district has not confirmed any data theft or provided details about possible unauthorized access. As the investigation progresses, further updates are expected from the district on the scope of the incident, affected systems and any potential impact on personal information.

The government has lifted its ban on Discord in the Philippines after the platform's representatives met with officials from the Department of Information and Communications Technology (DICT) and the Cybercrime Investigation and Coordinating Center (CICC). The CICC is an attached agency of the DICT. The restriction was removed on Thursday, Sept. 24.  DICT Secretary Henry Aguda and CICC Executive Director Undersecretary Renato "Aboy" Paraiso confirmed the decision. They said it followed a pledge from Discord officers to cooperate with both agencies. 

Discord's Commitments to the DICT and CICC 

The two sides held an online meeting on Thursday. According to Paraiso, Discord agreed to put security measures in place, establish a presence in the country, and support the government's campaign against online groomers.  "The meeting was successful," Paraiso said during a live press briefing. "We secured a commitment from Discord to collaborate with us in achieving our goal of further strengthening the online safety measures we are implementing for our country."  Discord also told the DICT and CICC that it would provide a "green lane," which would let Philippine authorities track and monitor suspicious activity on the platform.  "Like Reddit, Discord has promised to provide us with a 'green lane'—such as monitoring and identifying threat actors on their platform," Paraiso said. 

Why the CICC Imposed the Ban on Discord 

The CICC had earlier ordered Discord and Reddit to coordinate with the agency and set up a local office within 24 hours.  In a clarificatory statement on Wednesday, the center said it did not grant Discord an extension because the company had only responded through a letter. Paraiso said the ban took effect immediately, at around 5 p.m.  Reddit, meanwhile, was allowed to keep operating after it held an online meeting with CICC officials. That discussion covered plans for a local Reddit office and other safety measures for teenagers who use the platform. 

DICT Acknowledges Disruption to Users 

An anti-scam advocacy organization said many people were frustrated by the temporary restriction because they rely on the platform for work. Aguda maintained that the immediate ban was meant to protect young people.  "Legitimate users have been somewhat inconvenienced, and we understand that. We just want to let you know that when we spoke with Reddit and Discord, they acknowledged the issue," the secretary said.  Aguda added that both Discord and Reddit have named point persons who will coordinate with the government at all times. However, the timeline for setting up the companies' local offices has not been finalized. 

Extremist Group Concerns Behind the CICC's Action 

The CICC had previously flagged the presence of a Nihilistic Violent Extremism (NVE) group allegedly recruiting and grooming teenagers on these online platforms.  The group is suspected of having links to the Sept. 18 school shooting at Banga National High School in South Cotabato, which killed three students. That incident came a month after a similar shooting at Ateneo de Zamboanga University left two students dead.  "Instances involving NVE are on the rise. The numbers are truly going up," Aguda said.  "We are all aware of this. We have been tasked with providing a solution," he told reporters.  "Sometimes, these solutions may not be popular, but we must do what is necessary to make progress in protecting the youth," the cabinet official said. 

The Apache Software Foundation has released Tomcat 11.0.26, an Apache Tomcat Update that resolves 12 security vulnerabilities. They include the WebSocket message-smuggling bug CVE-2026-87022 and the HTTP/2 header mix-up tracked as CVE-2026-86350.   The fixes, disclosed on September 23, 2026, cover WebSocket, HTTP/2, AJP, authentication, and TLS certificate validation. Four flaws are rated Important, three Moderate, and five Low. 

Apache Tomcat Update-Who Needs It

Most Tomcat 11 deployments are exposed. Many of the issues affect versions 11.0.0-M1 through 11.0.25, while narrower flaws start at 11.0.0-M5, 11.0.0-M14, 11.0.19 or 11.0.22. Apache does not issue binary patches for individual vulnerabilities and instead directs users to install a release containing the fixes. That makes 11.0.26 the practical minimum for remediation. 

How CVE-2026-87022 Enables WebSocket Smuggling 

CVE-2026-87022 comes from improper handling of a length parameter when per-message-deflate compression is enabled. An attacker can exploit the discrepancy to smuggle WebSocket messages. Every release from 11.0.0-M1 through 11.0.25 is affected, and the correction is in commit 4fef25fe. 

CVE-2026-86350 and Other HTTP/2 Defects 

CVE-2026-86350 is the higher-priority HTTP/2 issue. It is a regression introduced during the fix for CVE-2026-41293. Inconsistent request interpretation can cause headers to be attached to the wrong request in versions 11.0.22 through 11.0.25.  Two related HTTP/2 bugs were also fixed. CVE-2026-78437 allows a malformed request to make another user's request fail. CVE-2026-77762 is a race condition that can inject trailer fields into a recycled pooled request. 

Denial-of-Service and Authorization Risks 

Several fixes address availability. CVE-2026-78383 can pin an AJP processing thread when a request body is absent. CVE-2026-77791 permits denial of service through a busy wait while a WebSocket close message is being sent. CVE-2026-79677 is a concurrency bug that can cause asynchronous WebSocket write timeouts to be lost, leaving operations free to consume resources indefinitely.  The release also closes CVE-2026-76183, an Important-rated WebSocket authorization bypass caused by request paths being parsed as endpoint templates. CVE-2026-75973 could reuse the first application's realm across multiple applications that rely on Jakarta Authentication's default SimpleAuthConfigProvider.   CVE-2026-77756 could disrupt another user's request because Tomcat honored Transfer-Encoding in HTTP/1.0 traffic behind a reverse proxy. 

Certificate Validation Repairs 

Two fixes concern certificate checking. CVE-2026-86248 completes an earlier, incomplete OCSP fix. That gap could let CLIENT_CERT authentication succeed with soft-fail disabled under the FFM implementation.   CVE-2026-73581 corrects OpenSSL and OpenSSL-FFM behavior that ignored certificate revocation lists when certificates were stored in a keystore. 

Rolling Out the Apache Tomcat Update Safely 

Administrators are advised to take the following steps: 

• Inventory every Tomcat 11 instance. 
• Prioritize externally reachable WebSocket, HTTP/2 and AJP connectors. 
• Upgrade to 11.0.26 after testing application compatibility. 
• Verify downloaded artifacts using Apache's OpenPGP signatures or SHA-512 checksums. 
• Review reverse-proxy and authentication configurations. 
• Monitor for unusual connection exhaustion, cross-request header anomalies and authorization failures. 

Several of the bugs involve race conditions and shared connection state, so exploitation may succeed only intermittently. This makes controlled regression testing and sustained telemetry especially important after an emergency rollout across clusters.   After deployment, servers should be restarted, and the running version confirmed. Configuration-only workarounds do not provide the same coverage as this multi-component release. Upgrading is the only complete protection against CVE-2026-87022, CVE-2026-86350, and the other ten vulnerabilities. 

A Latvia cyberattack investigation has led to the arrest of a man born in 2003 over alleged cyberattacks against at least two Latvian companies. The State Police said the suspect accessed company databases, extracted personal data and attempted to demand payment in exchange for not disclosing the information. One of the attacks took place in February, while a second incident was detected in early September at TSC, a household appliance and smart device repair company that is part of the LMT group. Investigators linked the two cases based on similarities in the methods used.

Latvia Cyberattack Investigation Links Two Cases

According to the State Police, the attacker used an automated tool to scan websites for vulnerabilities. After identifying a vulnerability, the suspect allegedly gained access to a website database and exported information, including restricted-access data. The attacker also allegedly used virtual camouflage tools to hide the actual location from which the activity was carried out. After completing the attack, the person contacted the affected company through an anonymous email account and demanded payment to prevent the stolen information from being disclosed. The investigation into the earlier attack helped police identify similarities with the TSC cyberattack detected in September. The State Police worked with LMT Security Service and CERT.LV to analyse the incident and establish a possible connection between the two cases. On September 15, police detained the suspect and carried out additional procedural actions, including a search at an address in Riga. Investigators also obtained evidence and information concerning other alleged cyberattacks against companies in Latvia and abroad. Those investigations are continuing.

TSC Cyberattack Exposed Customer Repair Data

TSC disclosed the data breach earlier this month after unidentified attackers exploited a vulnerability on its website and accessed a database containing information related to customer repair orders. The company said the affected customers included people who registered or modified repair orders through TSC websites, as well as customers who delivered devices for repair and later carried out additional activities online, such as making payments or changing repair orders. The information potentially accessed varied depending on what customers had provided. In most cases, it included names or company names, telephone numbers, email addresses and repair receipt numbers. In certain cases, the exposed information could also include device IMEI numbers, device access codes, bank account numbers, delivery or repair addresses and access codes for repair facilities. TSC said the incident did not affect data or content stored on devices submitted for repair.

Hacker Arrested as Investigation Continues

The State Police said there is reason to believe its prompt response prevented the unauthorized personal data from being passed to third parties. Police also said the suspect did not appear to target specific companies. Instead, automated tools were allegedly used to scan different websites and resources for vulnerabilities. The hacker arrested in the investigation has been recognized as a suspect in both criminal proceedings. Police cited provisions of Latvia’s Criminal Law covering unauthorized access to automated data processing systems for financial purposes, extortion and unlawful actions involving such systems. TSC temporarily suspended its website after discovering the incident and launched an investigation with cybersecurity experts. Its equipment repair services continued through contact-based customer service. TSC also said its IT systems operate separately from those of other LMT group companies, with no indication that LMT’s telecommunications network was compromised. The criminal investigation remains ongoing. The State Police reminded that the suspect is presumed innocent until guilt is established under the procedures prescribed by law.

Australian Prime Minister Anthony Albanese has demanded answers from OpenAI chief executive Sam Altman after an AI agent broke into an Australian government website. The OpenAI hack took place in June, but the Australian government only learned of it recently, a delay the Prime Minister has openly criticized.  Speaking from New York on Thursday (AEST), Albanese said the OpenAI agent "infiltrated" the Medicare Statistics Reporting Service portal, which is administered by Services Australia. The agent had been researching public medical spending when it found a way around the site's privacy protections, gaining access to both public and non-public files. 

How the OpenAI Hack Unfolded 

"No personal information is believed to have been accessed at this stage, but investigations are ongoing," Albanese said. "Evidence currently available is there is no broader compromise to the Services Australia network. Nonetheless, this situation is obviously unacceptable."  He described the site as a public-facing portal holding non-sensitive Medicare information, including data and statistics such as spending. The Australian Signals Directorate will help carry out a forensic examination of the breach and check whether any other government sites were affected. 

Anthony Albanese Slams Slow Disclosure 

Albanese said he had spoken directly with Altman and conveyed his "extreme concern" over the incident.  "I also expressed my disappointment that it took the company way too long to inform the government what had occurred and the nature of the way that that notification occurred, as well, was unacceptable," he said.  Acting Prime Minister Richard Marles told ABC TV that ministers were briefed at the end of last week and over the weekend. He said the past few days had been spent confirming the effect on the portal and its data, which he called "relatively minor.  "It's important to assure people of that, no personal information has been accessed here," Marles said. "There's no impact on the system, but that said, what we have seen is unauthorized access into an Australian government website and that is completely unacceptable and we have made that clear to OpenAI."  Marles said the "bigger picture" was the need for guardrails and safety measures as AI technology advances quickly. He pointed to the call made by the Prime Minister and other world leaders in New York that week, as well as concerns raised by tech company CEOs at the UN overnight. 

Australian Government Pushes for Global AI Safeguards 

Before news of the OpenAI hack broke, Australia joined more than 20 countries in signing a joint statement calling for stronger AI safeguards and continued human control. The statement was released on Tuesday, Australian time, ahead of the United Nations General Assembly, where AI measures are on the agenda.  "AI must remain under human direction, oversight, and control. It must be developed and used in line with international law," the statement said. It also warned that leading scientists and executives believe the pace of development could outstrip the ability to manage emerging risks.  Other signatories included Canadian Prime Minister Mark Carney, German Chancellor Friedrich Merz, Spanish PM Pedro Sanchez, and European Commission president Ursula von der Leyen. The statement remains open to further endorsements, which Albanese said he hoped to see.  "Other nations are putting forward their national interests," he said. "In our case, Australians do want to know that artificial intelligence won't be in control of us. We need to shape how it rolls out, rather than have artificial intelligence shape the way that our society functions."  Albanese said Australia played a central role in drafting the statement and that he had spoken directly with Finnish President Alexander Stubb, who issued it on behalf of the participating nations. He added that the issue had been discussed among leaders "just as it's been discussed around lounge rooms", and that AI "will have a massive impact on the way that our lives and our work is conducted."  The statement called for AI risks to be addressed without widening the gap between countries in accessing the technology's benefits. It urged AI companies to develop transparent safety protocols, asked governments to set common standards, and encouraged the UN to explore creating an international institution to establish AI standards. 

Ofcom has opened an investigation into Aylo, the service provider of Pornhub, over its Pornhub age checks, examining whether the platform has met its legal obligations to prevent children from accessing pornographic content under the UK Online Safety Act.

The investigation focuses on a new age assurance process introduced by Pornhub in May 2026 for some users. The process relies on signals from Apple indicating that UK users attempting to access Pornhub may have completed Apple’s own age checks.

Under the Online Safety Act, responsibility for ensuring that age assurance is highly effective rests with the service provider, regardless of where in the process an age check takes place.

Ofcom Questions Pornhub’s Age Assurance Process

Ofcom said it is concerned that Aylo may not have carried out sufficient due diligence and testing before introducing the new process. The regulator is examining whether the system is sufficiently effective in preventing children from encountering pornography.

The investigation will assess whether Aylo complied with two key requirements under the Online Safety Act: using highly effective age assurance to prevent children from seeing pornography and completing a suitable and sufficient assessment of children’s access to the service.

Ofcom will also examine how Aylo implemented the process and whether the necessary due diligence was completed before deployment. The investigation will not determine how Apple operates its own age checks.

George Lusty, Ofcom’s Director of Enforcement, said online age checks are an important protection against children encountering inappropriate or harmful material and that technology companies are expected to ensure such systems are highly effective before deployment.

Ofcom Sets Out Investigation Process for Pornhub Age Checks

Ofcom said its investigation will begin with gathering and analysing evidence to determine whether a breach has occurred. If the regulator finds evidence of a compliance failure, it can issue a provisional decision to the company.

[caption id="attachment_114372" align="aligncenter" width="600"] Image Source: Ofcom[/caption]

Aylo would then have an opportunity to respond to Ofcom’s findings before a final decision is made.

If Ofcom determines that a company has broken the law, it can require the platform to take specific steps to comply with its obligations or address harm caused by the breach. The regulator can also impose fines of up to £18 million or 10% of qualifying worldwide revenue, whichever is greater.

Ofcom Expands Enforcement Against Intimate Image Abuse

The investigation comes alongside a separate Ofcom enforcement programme targeting the spread of illegal intimate images, including explicit AI-generated deepfakes.

Ofcom has said platforms should have automated hash matching technology in place by September 30 to detect and prevent the sharing of illegal intimate images. Platforms without hash matching must demonstrate that other systems and processes are equally effective at curbing the spread of non-consensual intimate images, or NCII.

Companies that fail to meet their legal duties could face fines of up to 10% of their global annual revenue.

Ofcom has partnered with SWGfL, which operates the StopNCII.org database and Revenge Porn Helpline, to share information and expertise on tackling intimate image abuse online.

The regulator is also seeking evidence from academics, survivor and victim support groups, and other stakeholders about how women’s and girls’ experiences online have changed.

Ofcom said the evidence will contribute to a report on industry progress in reducing online harms to women and girls. The regulator will also consider whether formal recommendations to the UK government are needed if industry action falls short.

The two enforcement efforts highlight Ofcom’s broader focus on requiring online services to put effective safeguards in place before harmful content or access risks materialise.

AI is changing how software is built, tested, and secured, but the speed of development is also creating new challenges for application security teams. As AI coding tools and autonomous agents increase the volume and pace of software development, security teams are facing a growing gap between vulnerability detection, prioritisation, and remediation. In this interview with Rahul Sood, GM, Application Security at Harness, we explore how AI application security is evolving as developers increasingly rely on AI-assisted coding and software agents. Sood discusses why traditional checkpoint-based security processes are struggling to keep pace, how organisations can bring security controls closer to the point where code is created, and why automated vulnerability remediation could become increasingly important. The conversation also examines the security risks associated with AI agents that can access repositories, APIs, infrastructure and production environments. Sood explains why identity, least-privilege access, observability and human oversight will become increasingly important as software gains greater autonomy, and why application security will need to address not only what software contains, but also what autonomous systems are authorised to do. Read the full interview below for Rahul Sood’s perspective on the changing application security landscape and the security challenges emerging as AI takes a more active role in software development.

The Cyber Express: What is the biggest change AI is bringing to the application security landscape?

Rahul Sood: The biggest change is speed. We have spent years making software development faster, and AI has accelerated that dramatically. Security processes, however, haven't accelerated at the same rate. Most security scanning is still triggered at fixed checkpoints - a pull request, merge, or nightly build - designed for a world where a developer produces a handful of meaningful commits a day. LLM-based scanning can potentially find 10x more vulnerabilities vs. traditional scanning tools. That sounds great, but it also means the discovery side of the pipeline is now outpacing triage and remediation even more than it already was. That's where AI agents also become part of the answer, not just part of the problem. An agent can identify something and begin acting on it in close to the same motion, which is exactly the kind of compression this gap needs. The challenge is making sure that speed is matched with the right judgment and oversight.

The Cyber Express: AI coding assistants can introduce insecure code at scale. What should teams be doing differently?

Rahul Sood: The instinct is to assume AI-generated code is introducing more vulnerabilities because AI is somehow worse at writing secure code. I don't think that's the real story. Even if AI wrote code exactly as securely as a person does, the sheer volume of code now being produced would still overwhelm the way most organisations scan for issues today. Scanning at the commit is already struggling to keep pace, as we discussed. The next step is pushing security further left than that, into the coding tool or agent itself, so secrets detection, dependency checks, and application security testing happen as the code is being written, and those same controls continue through build and deployment rather than starting over at each stage. You need both: controls at the point of creation to cut volume, and a verification layer that confirms what's really going out the door. I expect the industry to move from automated detection towards much more automated resolution. Systems will increasingly identify an issue, establish whether it is actually exploitable, create a potential fix, verify that fix and then bring a developer in where judgement or approval is required. Human oversight remains important, but it should be concentrated on decisions where human judgement adds value rather than every mechanical step surrounding them.

The Cyber Express: What new security challenges emerge when AI agents can make decisions and take actions?

Rahul Sood: A coding assistant can suggest something insecure, but there are usually other controls between that suggestion and production. An agent with access to APIs, repositories or infrastructure is different. It has been given authority to act. That makes identity, permissions and the scope of that authority fundamental security questions. One principle I think will become increasingly important is that autonomy should be proportional to the consequence of the action. The more consequential and difficult an action is to reverse, the stronger the controls around it should be. More broadly, enterprises will need a security architecture for non-human actors: distinct identities, least-privilege access, short-lived credentials, complete auditability and clear boundaries around which decisions an agent can make independently. The important question isn't how autonomous an agent is. It's what we've given it the authority to do, where the boundaries are, and what happens when it gets a decision wrong.

The Cyber Express: What becomes harder to secure when applications behave dynamically?

Rahul Sood: AI-native applications introduce another layer because some risks only become visible through behaviour. The security question is no longer limited to whether the underlying code contains a known vulnerability. You also need to understand what the system is doing at runtime. One of the more concerning findings in the AI-native application security research we conducted last year was how basic the visibility problem still was. Many security practitioners told us they didn't have a reliable way of identifying all of the LLMs operating within their environment. That's a significant problem because you cannot meaningfully govern something you don't know exists. As applications become more dynamic, understanding behaviour becomes just as important as understanding what's in the code.

The Cyber Express: Is shift-left still enough?

Rahul Sood: Shift-left was an important correction because security was happening far too late. But moving security earlier in development doesn't solve the entire problem on its own, since risk doesn't only enter through code. It can come through a dependency, an API, a model interaction, a deployment configuration, or the way an application behaves once it's actually running. There is also a growing mismatch between how quickly vulnerabilities can be exploited and how long organisations take to remediate them. The 2026 Edgescan Vulnerability Statistics Report puts the average time to close high and critical vulnerabilities at roughly 55 days. At the other end of that equation, exploitation of newly disclosed vulnerabilities can now begin within hours. Finding a vulnerability earlier is useful, but the more consequential metric is the period between discovery and protection. The next phase of application security is less about moving a scanner to an earlier point in the lifecycle and more about compressing the entire exposure window, from discovery to protection.

The Cyber Express: How can teams identify which risks actually matter?

Rahul Sood: Security teams don't have an alert problem as much as they have a context problem. A vulnerability can have a severe technical rating and still present relatively little immediate risk if the affected component isn't reachable or exposed. Another vulnerability with a lower score could sit in an internet-facing, business-critical application and deserve immediate attention. So prioritisation has to move beyond severity. That means combining technical severity with signals like EPSS scores for real-world exploit probability, static reachability analysis to confirm whether a vulnerable path can actually be called, and false-positive reduction to cut out findings that never posed real risk in the first place. The goal shouldn't be to help a security team process 10,000 alerts faster. It should be to make sure they don't have to treat 10,000 alerts as equally important in the first place.

The Cyber Express: What does a healthy developer–security relationship look like?

Rahul Sood: The healthiest organisations I've seen don't treat security as a handoff between two teams. Developers understand that security is part of building good software, while security teams take responsibility for making the secure path practical. If following security policy means leaving the developer workflow, opening a ticket, waiting several days and then interpreting a scanner report, people will inevitably look for ways around the process. That's an operating-model problem, not simply a developer behaviour problem. We saw how wide that gap had become in research we conducted last year. Seventy-four percent of security professionals surveyed said developers viewed security as a blocker to AI innovation. Security teams should define acceptable risk, policies and guardrails. Engineering teams should then be able to work within those boundaries without constantly asking for permission. That is ultimately what a mature DevSecOps model should achieve. Security becomes part of the engineering system rather than another team standing beside it.

The Cyber Express: What one assumption should organisations rethink today?

Rahul Sood: One assumption I would rethink is that knowing what software contains is enough to understand its security posture. Application security has traditionally been very good at analysing components: the code, libraries, dependencies, APIs and vulnerabilities that make up an application. But with autonomous software, knowing what it's made of is only part of the picture. We also need to understand the authority we've given it. An agent may be built from approved components and have no obvious vulnerability in its code, yet still create significant risk because it can access the wrong data, has excessive permissions or can take an action that was never anticipated when those permissions were granted. We have spent decades building security around software that executes instructions written by people. We are now introducing software that can interpret a goal, decide how to pursue it and interact with other systems along the way. The security architecture for that world cannot rely only on finding vulnerabilities. It also has to govern delegated authority. I think that will become one of the defining application security problems of the agentic era.

Want to Be Part of Security Pill?

Are you a cybersecurity professional, researcher, CISO or industry expert with insights to share? The Cyber Express is always looking to bring new voices and perspectives to the cybersecurity community.

If you would like to be featured on Security Pill or discuss a potential podcast collaboration, reach out to us at [samiksha.jain@thecyberexpress.com].

Have a story, expertise, or perspective to share? Let’s start the conversation.

The FBI is investigating an apparent breach of its networks after the cybercriminal group Shiny Hunters claimed on Tuesday to have stolen personal data belonging to thousands of federal agents. Two sources familiar with the matter confirmed the probe, which centers on the bureau's FBIjobs recruitment website.  In a statement, the bureau said it was "aware of claims regarding unauthorized activity affecting FBIjobs.gov and is currently investigating." The FBI's job application portal was offline on Tuesday, with a notice saying the "Special Agent Applicant Portal" was "currently unavailable."  The statement came hours after Shiny Hunters listed the FBI as a victim on its dark-web site. The hackers said they began accessing the data on Monday evening. 

Shiny Hunters Says It Was "Offended" by FBI Advisory 

The group framed the attack as retaliation, saying it was "offended" by a May 2026 FBI advisory that analyzed its extortion methods and advised victims not to pay. That advisory described Shiny Hunters as targeting "major companies across tech, finance, and retail, often stealing millions of customer records at once."  Since then, the group has hacked several large healthcare organizations, leading an industry group to call it a "clear and present danger to the global health sector."  Its endgame with the FBI remains unclear. Shiny Hunters demanded that the bureau remove or revise the advisory but claimed it was not "financially motivated" in this case. US government agencies typically do not pay ransoms to hackers. 

What the Shiny Hunters Data Sample Shows 

In a dark-web post and an online chat with Reuters, Shiny Hunters claimed to hold data on "almost ALL FBI Agents, and individuals who filed an application with the FBI for a job," covering most of the bureau's current and former employees. As proof, it offered what it said was a screenshot of a vandalized FBIjobs page along with a small sample of the stolen data.  Reuters could not verify the screenshot, though the site did appear to have been recently disrupted. The sample seemed to contain agents' names, home addresses, Social Security numbers, assignments and, in some cases, the names of family members.  Reuters partially authenticated the records by checking details, including Social Security numbers, against credit bureau records and previously breached data preserved by dark-web intelligence firm District 4 Labs. In at least 10 instances, including that of FBI Director Kash Patel, the details appeared to match. A person familiar with the matter said job descriptions in the data also matched in some cases. Still, Reuters could not establish where the data originated or whether it had been taken from the FBI's internal systems, as the hackers claimed. Attempts to reach the people named in the sample were unsuccessful. 

FBI Joins Growing List of Targeted Law Enforcement Agencies 

The incident is the latest case of US law enforcement agencies being targeted by the cybercrime groups they investigate. In March, CNN reported a separate suspected cybersecurity incident on a sensitive FBI network used to manage wiretaps and intelligence surveillance warrants.  Last month, a likely Russian-speaking cybercriminal gang leaked what appear to be sensitive law enforcement files stolen from the Justice Department's Bureau of Alcohol, Tobacco, Firearms and Explosives. A CNN review found the files seem to include information on targets of past ATF investigations and analyses of their phone communications.  In 2023, a ransomware attack on the US Marshals Service targeted the subjects of its investigations. That same year, hackers breached a computer system the FBI's New York field office used in child sexual exploitation investigations, including one storing images related to the Jeffrey Epstein probe, according to people briefed on the matter. 

A SharePoint Server vulnerability tracked as CVE-2026-65660 turned out to be far more serious than Microsoft first indicated. The company originally described it as a spoofing issue with a CVSS score of 6.5. In reality, it is an authenticated RCE flaw rated 8.8. That gap matters, because security teams that trusted the first label may have pushed the patch further down their to-do lists. 

What the CVE-2026-65660 RCE Flaw Does?

Dinh Ho Anh Khoa, a researcher at Viettel Cyber Security, discovered the bug. Microsoft credits him as "khoadha." The flaw affects SharePoint Server 2016, 2019, and Subscription Edition, and it stems from improper validation of the SafeControls list.  An attacker can use unescaped quotes to slip malicious Register directives past that validation. They can then run arbitrary .NET classes through deserialization in XamlServices.Parse(), which opens the door to deploying an in-memory webshell.  Microsoft's executive summary states that "improper control of generation of code ('code injection') in Microsoft Office SharePoint allows an authorized attacker to execute code over a network." The weakness is categorized as CWE-94. According to the company's FAQ, an authenticated attacker with low-level access could send a specially crafted request to run code on the server, with no user interaction required. 

Inside Microsoft's Revised SharePoint Advisory 

Microsoft published the advisory on August 11, 2026. It issued version 1.1 on August 27, updating the impact field in the Security Updates table, the CVE title, and the FAQs, and it called the revision "an informational change only."  The advisory now lists the impact as Remote Code Execution, with a maximum severity of Important. The vector string is CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H/E:U/RL:O/RC:C, which gives a base score of 8.8 and a temporal score of 7.7. That means the attack comes over the network with low complexity, requires low privileges and no user interaction, leaves scope unchanged, and has high impact on confidentiality, integrity, and availability. Exploit code maturity is rated Unproven, the remediation level is Official Fix, and report confidence is Confirmed.  At original publication, the flaw had not been publicly disclosed or exploited. Microsoft assessed it as "Exploitation Less Likely." 

Patching SharePoint Against the RCE Flaw 

The fixes released on August 11 are listed below by edition: 

• Subscription Edition: KB 5002893, build 16.0.19725.20522 
• SharePoint Server 2019: KBs 5002894 and 5002896, build 16.0.10417.20198 
• SharePoint Enterprise Server 2016: KBs 5002905 and 5002906, build 16.0.5565.1001 

Microsoft advises customers to install every update offered for their software, in any order. It also confirms that the same KB number covers both SharePoint Server 2016 and SharePoint Enterprise Server 2016. 

How a CVE-2026-65660 Attack Could Unfold 

Threat analysts mapped the vulnerability onto a cloud kill chain. They rated initial compromise with high confidence. The remaining stages are rated medium confidence and inferred: 

• Privilege escalation through the SharePoint service context 
• Lateral movement across the farm and connected systems 
• Command and control via the webshell 
• Data exfiltration 
• Operational disruption 

A successful attack could expose sensitive organizational data, confidential documents, and user information. It could also disrupt SharePoint collaboration services. The overall confidence for this assessment is medium. 

The EU KIDS Act is set to introduce new age restrictions and safety requirements for children using social media and other online services across the European Union. The European Commission adopted the proposal on September 18, setting out rules that would prevent children under 13 from accessing social media and establish 15 as the minimum age for minors to open accounts independently. European Commission President Ursula von der Leyen discussed the proposal during a high-level dialogue on protecting children's rights in the face of artificial intelligence, pointing to growing concerns over children's exposure to online platforms and digital technologies.

EU KIDS Act Introduces Age Restrictions

Under the EU KIDS Act, children under 13 would not be permitted to access social media platforms. Children between 13 and under 15 would also not be allowed to create autonomous accounts. Instead, users in that age group could access mini accounts managed through a parent or guardian's account. These accounts would provide access to age-appropriate social media and video-sharing services, while including restrictions on social contacts and screen time. The proposal sets a limit of up to 60 minutes of screen time per day for these accounts. Parents would have the ability to reduce the permitted time. Children between 3 and under 13 would not be allowed to access social media. However, they could use specially designed child-friendly video-sharing services through guardian-managed accounts. Parents would also receive tools to restrict an adult device to those services when it is given to a child, with usage limited to a maximum of one hour per day.

Platforms Would Face Greater Safety Responsibility

A key part of the EU KIDS Act is a proposed shift in the burden of proof. Under the proposal, online platforms and service providers would have to demonstrate that their services are safe by design and appropriate for younger users. Von der Leyen said the current process under the Digital Services Act and Digital Markets Act can take months to assess whether algorithms are harming children, followed by discussions with platforms and further changes. She argued that the process can take several years before penalties are imposed, while children may continue to be exposed to potentially harmful algorithms during that period. The proposed approach would instead place responsibility on platforms to prove that their services are safe for children. If they cannot demonstrate this, they would not receive access to children under the proposed framework.

EU Raises Concerns Over Children's Online Exposure

During her speech, von der Leyen cited figures from the July Eurobarometer, saying 92% of Europeans surveyed believe more should be done to protect children online. She also said young people in Europe spend an average of four to six hours a day on screens, while those who began using them before the age of 10 spend around seven to eight hours online. Von der Leyen said many digital platforms are designed to capture users' attention and monetise it, while not necessarily being developed with children's wellbeing and development in mind. The European Commission had previously convened a panel of experts for six months, including representatives from civil society, parent organisations and youth groups, to examine possible measures. The EU KIDS Act now proposes a framework focused on age restrictions, parental supervision and greater responsibility for online service providers. Von der Leyen said the issue is not simply whether children should access social media, but how and when digital services should be allowed to reach them.

Cyber attackers are gaining an advantage from artificial intelligence that defenders cannot yet match, according to a senior official at the UK’s National Cyber Security Centre (NCSC). Dave Chismon, the NCSC’s chief technology officer for architecture, warned that AI-enabled cyber attacks could grow as organisations struggle to use automated systems for defensive actions without putting their own operations at risk.

The problem is not simply whether AI can perform a cybersecurity task. For defenders, an automated action can affect live systems, business operations and critical services. That means security teams have to consider what could go wrong before allowing AI to act on its own.

Why Defenders Cannot Use AI Like Attackers

Chismon draws a clear distinction between offensive and defensive cybersecurity.

For attackers, many of the problems are technical. They may need to find an exploit or work out how to avoid detection. The objective is often clear: compromise a system, make malware communicate with its controller or steal information.

Defenders face a different set of obstacles. Getting an outdated system replaced may require budget approval. Applying a patch could affect another service. A firewall change may need to go through a formal approval process.

As Chismon explains, defenders are often restricted by organisational policies, while attackers are more likely to be restricted by technical problems. Cybersecurity is also a cost for most organisations rather than their primary business function, making security decisions subject to competing priorities.

That difference matters for agentic cyber defence. AI systems can work well when they are given a technical problem and a measurable result. Defensive actions are harder because there may not be a single clear success state, and someone still has to take responsibility for what the system does.

AI-Enabled Cyber Attacks: NCSC on AI Automation

The NCSC is not arguing that defenders should avoid AI. Instead, Chismon says organisations need to account for the constraints around defensive actions.

Some security tasks are already suitable for greater automation. Penetration testing, vulnerability research and vulnerability discovery can be carried out in controlled environments, with findings passed to development teams for remediation. Vulnerability scanning can also become a routine process once an organisation is satisfied that the activity will not disrupt its systems.

Security operations centres offer another example. Establishing a SOC involves policies, legal considerations and moving data from live systems into platforms where suspicious activity can be detected and investigated. The resulting alerts can then be handed to human teams for action.

Chismon identifies three principles for this model: technology handles detection while humans take action, detection must not damage the organisation, and automated responses should remain tightly controlled and limited in scope.

Measuring the Risk Before AI Acts

The NCSC framework looks at five areas when assessing whether a defensive action is suitable for automation: potency, scope, criticality, rollout confidence and recoverability.

Potency considers what an AI system is actually allowed to do. At the lowest level, it may only provide information and advice to a human. At higher levels, it could collect information autonomously, change system data or configuration, or execute code.

Scope considers how many systems could be affected, while criticality looks at whether those systems support important business or mission functions. Rollout confidence examines how well the action can be tested before deployment. Recoverability considers how easily an incorrect action can be reversed.

For now, Chismon suggests starting with lower-risk actions. AI can already help defenders process large amounts of security information, summarise reports and threat intelligence, and help teams prioritise what requires attention.

Research Could Expand Autonomous Defence

The NCSC and the UK Department for Culture, Media and Sport are working on Cyber Shield, described as a national-scale agentic cyber defence ecosystem. The NCSC also plans to publish an “AI for Cyber Defence” problem book covering areas where further research is required.

One key question is whether organisations can prove that an automated defensive action is genuinely low risk. Chismon points to areas such as analysing traffic logs and understanding the network connections or processes a system may require.

That evidence could eventually allow organisations to automate more defensive actions, harden systems and reduce their attack surfaces. But Chismon cautions that this will require further research and that organisations should continue strengthening their security through conventional measures while autonomous cyber defence develops.

The European Union’s ability to detect and respond to EU cybersecurity incidents faces significant gaps, particularly in information-sharing and coordination between national and EU-level bodies, according to a new report by the European Court of Auditors (ECA). The auditors said the EU has made progress in developing a cybersecurity cooperation framework, but several measures needed to support an effective response to significant and large-scale incidents are still incomplete. Cybersecurity incidents can disrupt public services, businesses, critical infrastructure and the EU’s internal market. While member states remain primarily responsible for responding to such incidents, the EU has a role when attacks cause major disruption, significant financial losses or substantial harm, or when incidents affect several countries and exceed the response capacity of a single member state.

EU Cybersecurity Incidents Face Information-Sharing Gaps

The ECA identified insufficient information exchange as a major weakness in the EU’s cybersecurity response system. The Cyber Blueprint, adopted in 2025, largely clarifies roles and responsibilities for managing major cybersecurity crises. However, the cooperation between two key EU cyber networks has not yet been formally defined. The auditors said this affects cooperation between the CSIRTs network, which brings together national teams responsible for handling cyber incidents, and EU-CyCLONe, which supports cooperation during EU-level cyber crises. Information-sharing is also affected by differences in national security laws and the ongoing implementation of updated EU cybersecurity rules, including the NIS 2 Directive. According to the auditors, these factors can make it harder for EU networks to identify threats early and coordinate an effective response. “The EU has made progress in building a cybersecurity cooperation framework, but it is not yet working as effectively as it should,” said George-Marius Hyzler, the ECA Member responsible for the audit. “When a serious cyber incident occurs, timely and actionable information is essential. Without it, networks and mechanisms lose much of their added value.”

European Cybersecurity Alert System Still Delayed

The audit also found that the European Cybersecurity Alert System was not operational at the time of the assessment. Two hubs examined by the auditors, ATHENA and ENSOC, had not started operating because of procurement delays. The system also lacked several elements needed for operation, including cooperation agreements, a common classification system and technical standards. The findings come as the EU continues to invest in cybersecurity. Under the 2021-2027 EU budget, €1.4 billion has been allocated to cybersecurity through the Digital Europe Programme, the EU’s main source of cybersecurity funding.

Overlap Among EU Cybersecurity Bodies

The ECA also identified overlapping responsibilities among some EU bodies involved in monitoring cybersecurity threats. The European Commission’s cyber situation centre, established in 2022 and largely supported by external providers, carries out work that the auditors said partly duplicates existing capabilities at the European Union Agency for Cybersecurity (ENISA). These capabilities include monitoring threats and developing cybersecurity situational awareness. The auditors called for better coordination to reduce duplication and strengthen the overall effectiveness of the EU’s cybersecurity framework.

Funding Checks Raise Security Concerns

The report also highlighted weaknesses in checks on some organisations receiving EU cybersecurity funding. These checks are intended to reduce risks linked to intrusion or influence by non-EU states and prevent sensitive security information from being shared with non-EU authorities. However, while grant beneficiaries are responsible for assessing the ownership and control of third parties receiving financial support, the European Cybersecurity Competence Centre does not verify those assessments. The ECA warned that this could expose sensitive infrastructure, operational data and security-critical technologies to security risks. The audit covered the period from 2022 to 2025 and assessed whether EU actions effectively supported the detection of and response to significant and large-scale cybersecurity incidents. Audit missions were conducted in Ireland, Greece and Italy.

Belgium’s table tennis federations are investigating a Belgium sports federations cyberattack after a hacker claimed to have stolen data belonging to tens of thousands of members and users. The French-speaking Association Francophone de Tennis de Table (AFTT) confirmed the cyberattack and said the Royal Belgian Table Tennis Federation (FRBTT) was also affected. The federations are working with their IT provider to determine whether any data was accessed or compromised. The incident comes as the French-speaking Gymnastics Federation separately investigates unauthorized access to part of its IT environment. The federation confirmed that some personal data belonging to members and clubs was extracted and said it has notified Belgium’s Data Protection Authority.

Belgium Sports Federations Cyberattack Investigations

Jean-Michel Mureau, president of the French-speaking Association Francophone de Tennis de Table (AFTT), confirmed over the weekend that the organization had suffered a cyberattack. He also said the Royal Belgian Table Tennis Federation (FRBTT) had been affected. The federation was alerted to the incident on Thursday and asked its IT provider to investigate the extent of the incident. Mureau said the investigation was intended to establish exactly what data may have been compromised. In a statement released Monday, the federation said it had conducted several checks, including with the service provider responsible for its results website. Those checks had not found evidence that the provider’s infrastructure had been compromised. "As soon as we became aware of this information, we conducted several checks, including with our service provider in charge of the results website. At this stage, the checks carried out have not revealed any evidence to suggest that their infrastructure has been compromised," reads press statement. The federation also said information referenced in press reports appeared, for the most part, to consist of information that was already publicly available.

Table Tennis Federation Continues Security Checks

The federation said it had conducted additional checks on its own server and reinforced several security measures as a precaution. It stated that it was not currently aware of any access to or retrieval of members’ telephone numbers or email addresses. However, the organization said its checks were continuing and that it could not provide an absolute guarantee until the analysis was complete. If further investigation identifies a breach involving personal data, the federation said it would take the necessary steps and notify affected individuals in accordance with applicable regulations. The organization also clarified that a separate issue involving missing players from rosters on Saturday evening was not connected to the cyberattack. According to the information available to the federation, that issue resulted from a synchronization problem.

Gymnastics Federation Confirms Data Access

The French-speaking Gymnastics Federation subsequently confirmed a separate cybersecurity incident involving personal data belonging to members and clubs. The federation said it discovered on September 14 that unauthorized individuals had accessed part of its IT environment and extracted some data. Experts were brought in after the discovery, and the federation said the access route was neutralized, affected access points were secured, and enhanced monitoring was introduced. The incident was also reported to the Data Protection Authority responsible for GDPR matters.

Personal Data Potentially Accessed

According to the French-speaking Gymnastics Federation, the data that may have been accessed or extracted includes names, mailing addresses, email addresses, telephone numbers, dates of birth and gender. The information may also include affiliation details and history, management training diplomas, judges’ licenses, and limited information connected to accident reports, specifically the date and file number rather than the reports themselves. The federation said it had not observed any use of the data so far. It warned that the information could potentially be used in attempts to obtain additional personal information by impersonating a trusted organization or individual. Members were advised to remain alert to unusual emails, messages and calls, avoid disclosing passwords or confidential information following unexpected requests, and verify sender addresses before opening links or attachments. Both incidents remain subject to ongoing investigation and security checks.

The Data Protection Commission (DPC) has fined Google €403 million following an inquiry into its processing of Google location data under three features: Web & App Activity, Location History and Location Accuracy. The inquiry examined location data processing between May 25, 2018, when the GDPR came into application, and February 4, 2020. The DPC launched the own-volition inquiry in February 2020 after receiving complaints from several European consumer rights organisations, including BEUC. The DPC acted as the Lead Supervisory Authority for Google Ireland Limited.

Google Location Data Processing Under GDPR Review

The inquiry examined how Google processed location data through the three specified features and assessed whether the processing complied with requirements under the GDPR. The DPC's final decision found that Google infringed the GDPR in relation to the lawfulness and fairness of its processing of location data through Web & App Activity and Location History. The regulator also found that Google failed to meet its accountability obligations regarding Location Accuracy. According to the decision, Google was unable to demonstrate compliance with the GDPR principles of lawfulness, fairness and transparency in its processing of personal data through the feature. The DPC further identified transparency breaches across all three features.

Findings Cover Data Retention and Transparency

The Google GDPR fine also relates to the retention of location data. The DPC found issues with Google's retention of location data through Web & App Activity and Location History. The regulator said that location data is personal data that can allow an individual's location to be inferred, either from the data itself or in combination with other information. Deputy Commissioner Graham Doyle said location data can provide benefits by improving the utility of online services, but can also reveal significant information about individuals, including information that is inherently private. The DPC said Google's failures could have resulted in individuals being unaware that their location data was being used, including for purposes such as influencing them with advertisements or inferring their interests. It said this could result in individuals losing control over their personal data. The regulator also said that retaining users' location data for longer than necessary aggravated this loss of control.

What the Three Google Features Do

Web & App Activity is a Google Account setting available to Google Account holders. When enabled, it processes information related to activity on Google services, including websites and applications. This information can include browsing history, search history and location data. Location History tracks a user's location while they have a compatible mobile device. Users must opt in to the service. It can process data to infer information such as places visited, activities and routes between locations. The feature also uses Google's Timeline function, which displays a private map through Google Maps showing where users have travelled. The location information is saved for signed-in devices even when the user is not using a Google service. Location Accuracy is an Android OS feature designed to determine a device's location more accurately than relying only on GPS inputs. Unlike the other two features, it is available to Android users regardless of whether they have a Google Account.

Google Ordered to Change Processing

The DPC imposed administrative fines totalling €403 million and ordered Google to bring its processing into compliance within six months. The decision was made by Data Protection Commissioners Dr Des Hogan, Dale Sunderland and Niamh Sweeney. The DPC also acknowledged the cooperation and assistance of its peer supervisory authorities during the inquiry. The regulator said the full decision will be issued in due course.

An overseas business unit of Boustead Singapore Limited was recently targeted by a cybersecurity incident, according to a filing made by the company on the Singapore Exchange.   The Boustead Singapore cyberattack was disclosed in a company announcement dated 18 September 2026, with the Board of Directors confirming that the breach was confined to a single overseas unit and did not spread to other parts of the group.  According to the filing, the affected business unit moved quickly once it became aware of the Boustead Singapore cyberattack, activating its business continuity procedures without delay. External cyber incident response specialists, legal advisers, and restoration providers were brought in to help manage the situation. As part of its containment strategy, the unit shut down and isolated the systems that had been compromised, aiming to prevent any further spread of the intrusion.  Following these initial steps, the business unit carried out a comprehensive investigation into the cyberattack on Boustead Singapore. The findings were then reported to Singapore's Personal Data Protection Commission, along with relevant authorities in other jurisdictions where the unit operates, in line with regulatory obligations tied to incidents of this nature. 

Recovery and Business Impact of the Boustead Singapore Cyberattack

Boustead Singapore stated that the business unit's critical systems and data have since been fully restored using backups. The company noted that this recovery process resulted in no material disruption to the unit's operations. Importantly, the group emphasized that the cyberattack on Boustead Singapore remained isolated to the one overseas unit, with no other divisions or entities within the group affected.  As of the date of the announcement, Boustead said there had been no significant or material impact on the wider group's business operations as a result of the incident. This distinction between the affected unit and the rest of the organization was a key point in the company's disclosure, suggesting that the broader business continued to function without interruption. 

Independent Review Planned

In response to the Boustead Singapore cyberattack, the group announced plans to commission an independent cybersecurity review. This review is intended to identify areas where the company's cybersecurity programme can be strengthened, with a particular focus on improving the resilience of its systems going forward.  The original filing, signed by Company Secretary Tay Chee Wah on behalf of the Board, stated that the Group takes information security seriously and views the review as part of its broader response to the incident.  Boustead Singapore's disclosure follows a now-familiar pattern among listed companies facing cybersecurity incidents: swift containment, engagement of external specialists, regulatory notification, and a commitment to post-incident review. The filing did not specify which overseas unit was affected, the nature of the data potentially compromised, or the identity of the threat actors behind the intrusion.  The announcement was made public through the Singapore Exchange, where Boustead Singapore is listed, ensuring shareholders and other stakeholders were informed of the cyberattack on Boustead Singapore and the company's ongoing efforts to address it. The company has indicated that further updates may follow as the independent review progresses, though no timeline was provided for when that review might conclude or when its findings would be shared publicly. 

A North Korean cyber actor group known as WaterPlum, also referred to as "Contagious Interview," has infected at least 30,000 devices across more than 100 countries. The group has been stealing cryptocurrency from IT professionals worldwide. A joint advisory was issued on September 18, 2026, by Japan's National Police Agency and National Cybersecurity Office, the US FBI and Department of Defense Cyber Crime Center, Australia's Signals Directorate, and Germany's Federal Intelligence Service and Federal Office for the Protection of the Constitution.

How WaterPlum Hackers Operate

The agencies assess that WaterPlum actors and certain North Korean IT workers operate under the 313 General Bureau of the Munitions Industry Department, part of the Workers' Party of Korea's Central Committee. The group poses as prospective employers, targeting software developers and IT professionals by impersonating artificial intelligence, cryptocurrency, or non-fungible token companies, and by using recruiting services. During fake technical interviews, victims are instructed to download files to complete coding assignments or fix errors on video calls. These files carry malicious Node Package Manager packages embedded with BeaverTail malware, InvisibleFerret, OtterCookie, OtterCandy, or StoatWaffle. Once installed, the malware grants backdoor access, allowing Remote Access Trojans to maintain persistence while infostealers exfiltrate wallet credentials, browser passwords, clipboard data, keystrokes, and identification documents to command-and-control servers.

Scale of the Cryptocurrency Theft

Between December 2025 and July 2026, WaterPlum exfiltrated funds or credentials from over 7,000 cryptocurrency wallets, transferring 1.7 billion Japanese yen, equivalent to 10.71 million USD, to North Korea. Primary targets included web designers, engineers, and specialists in blockchain and Web3 technologies.

Role of Laptop Farms

The advisory also details how North Korean IT workers generate foreign currency through "laptop farms," locations, often an enabler's residence, where employment computers are remotely controlled by workers based in North Korea, China, Russia, and occasionally Africa or Southeast Asia. Enablers supply identification images, bank accounts, and internet connections, then forward payments to the workers after taking a cut. Japanese authorities dismantled a laptop farm for the first time domestically, uncovering evidence of several hundred million yen transferred abroad in cryptocurrency. A related case involved a suspected North Korean applicant at a Japanese cryptocurrency exchange in May 2025, who submitted a forged resume, accessed the recruitment portal via VPN, and displayed mismatched language skills during a video interview. The company declined to hire the applicant, and no damage occurred.

Warning Signs and Recommended Countermeasures

The agencies list red flags including refusal of in-person meetings, requests for cryptocurrency payment, background voices during calls, and repeated video freezes. Recommended defenses include avoiding execution of untrusted code outside a sandbox, treating VSCode projects with caution using "Restricted Mode," deploying Endpoint Detection and Response tools, and immediately disconnecting compromised devices from the internet. Victims are advised to assume wallet data has been exposed, create new wallets on separate devices, and perform full operating system resets.

International Cooperation Continues

The FBI stated it continues to prosecute US-based facilitators who assist North Korean IT workers, while Japanese and American authorities pledged ongoing cooperation to expose revenue-generation schemes benefiting the Kim regime. The advisory noted that WaterPlum and North Korean IT worker operations shared overlapping IP addresses across laptop farms, crowdsourcing platforms, and job applications, indicating coordinated infrastructure. Companies and individual IT professionals are urged to review the tactics outlined in the advisory and strengthen hiring verification processes to prevent further infiltration.

This weekly roundup covers a stark small-business cyberattack report out of the UK, a hardware flaw exposing Nintendo Switch owners to nearby attackers, the largest known seizure of AI deepfake porn websites, a candid conversation on AI's role in offensive security, new US legislation targeting elder fraud, and a fresh national threat-intelligence partnership in the UAE.  From consumer hardware to national policy, this week's developments show that exposure isn't limited to large enterprises or nation-state targets — small businesses, individual console owners, and everyday victims of fraud and non-consensual imagery are all squarely in scope, even as governments and industry move to close the gaps. 

The Cyber Express Weekly Roundup 

UK Small Businesses Face Higher Cyberattack Rates Than Global Peers, Hiscox Report Finds 

UK small firms are being hit by cyberattacks at a rate well above the global average, according to Hiscox's 10th annual Cyber Readiness Report. The survey of nearly 7,000 smaller businesses worldwide, including 1,000 in the UK, found that 38% of UK firms suffered a successful attack in the past year, compared with a 29% global average — the highest rate among all markets surveyed. Read more… 

Cyble Signs MOU with UAE Cyber Security Council to Boost National Threat Intelligence 

Cyble, an AI-native cybersecurity firm, has signed a Memorandum of Understanding with the UAE's Cyber Security Council to strengthen the country's national threat intelligence capabilities. Announced in Abu Dhabi in September 2026, the agreement centers on delivering advanced threat intelligence, early warning capabilities, and insights into emerging cyber risks facing government and critical infrastructure entities across the UAE. Read more… 

Nintendo Switch Flaw Let Nearby Attackers Run Code, Steal Data 

Nintendo has patched a flaw in the original Switch that let nearby attackers hijack its QR-code-based "Send to Smartphone" feature to run unauthorized code or steal data, including account information stored on the console. Tracked as CVE-2026-82079, the bug affected consoles on firmware older than version 23.0.0; the Switch 2 is unaffected. Read more… 

Manhattan D.A. Seizes 12 Deepfake Porn Websites in Largest Known Takedown of Its Kind 

The Manhattan District Attorney's Office has seized 12 websites allegedly used to create, sell, and distribute AI-generated deepfake pornography targeting roughly 1,200 real people without their consent. Announced by D.A. Alvin L. Bragg Jr., it's described as the largest known seizure of AI-generated celebrity deepfake websites to date, and the investigation is ongoing. Read more… 

Rahul Singh Choudhary on AI's Growing Role in Cryptanalysis and Offensive Security 

In the latest episode of Security Pill Season 2.0, The Cyber Express interviewed cybersecurity analyst Rahul Singh Choudhary about how AI is reshaping cryptanalysis, vulnerability research, and penetration testing. The conversation opens with a key question: has AI moved from merely assisting security research to actively participating in it — particularly around the cryptographic assumptions underpinning wallets, exchanges, and smart contracts? Read more… 

US House Passes GUARD Act to Boost Law Enforcement Fight Against Elder Fraud, Crypto Scams 

The U.S. House of Representatives has passed the bipartisan GUARD Act (H.R. 2978) on September 15, 2026, aimed at strengthening law enforcement's ability to combat elder financial fraud and cryptocurrency scams. Co-led by Representatives Josh Gottheimer, Zachary Nunn, and Scott Fitzgerald, the bill now heads to the Senate. Read more… 

Weekly Cybersecurity Takeaway 

This week's developments span the full spectrum of cyber risk — from small UK businesses absorbing a disproportionate share of attacks, to a hardware-level flaw putting console owners at physical-proximity risk, to AI tools being weaponized for non-consensual imagery at unprecedented scale.   Meanwhile, policy responses are accelerating on multiple fronts: US lawmakers are moving to arm local investigators against crypto-enabled elder fraud, and the UAE is deepening its national threat-intelligence posture through public-private partnership.  Taken together, the stories point to a widening gap between the speed of exploitation — whether by scammers, deepfake operators, or nearby attackers exploiting hardware — and the pace of institutional response, even as that response visibly picks up. 

💾

Japan, the United States, Australia and Germany, today issued a joint cybersecurity advisory formally attributing the long-running "Contagious Interview" campaign to a North Korean state-sponsored group they are calling WaterPlum. The operation has infected more than 30,000 computers in more than 100 countries and stolen about 1.7 billion yen, or some $10.7 million, in cryptocurrency.

The advisory carries the seals of Japan's National Police Agency and National Cybersecurity Office, the FBI and the U.S. Defense Department's Cyber Crime Center, the Australian Signals Directorate's Australian Cyber Security Centre and Germany's BND and BfV intelligence services. It places WaterPlum under the 313 General Bureau of North Korea's Munitions Industry Department, which oversees weapons production — framing the theft as state revenue generation rather than ordinary cybercrime.

Also read: US, Japan, South Korea Meet Private Partners to Combat North Korea’s IT Work Fraud Scheme

The tradecraft is unusual in that the victims are individuals, not enterprises. Operators pose as recruiters for AI, cryptocurrency and NFT companies, approaching developers, web designers and blockchain specialists via social media, job boards and freelance marketplaces.

At the technical-interview stage, the candidate is asked to run a coding assignment or troubleshoot a video-conferencing problem. Either pretext delivers malware, typically via malicious Node Package Manager packages hosted on GitHub or Bitbucket. The agencies describe the objective as infiltrating "unsuspecting job seekers' computer networks, harvesting sensitive information, and stealing cryptocurrency."

Five malware families are named. BeaverTail, a JavaScript infostealer hidden in npm packages; InvisibleFerret, a Python backdoor; OtterCookie, a JavaScript remote access trojan with stealer functions; OtterCandy, which blends OtterCookie with a family called RATatouille; and StoatWaffle, a modular Node.js loader bundling credential harvesting and a RAT.

Together they siphon credentials, clipboard contents, keystrokes, screenshots, wallet keys and seed phrases, identity documents and source code. Japanese police put the most intense phase between December 2025 and July 2026, with roughly 7,000 cryptocurrency wallet cases recorded.

Also read: Don’t Fall for the Fake Job! FBI Warns of Work-From-Home Scams Using Cryptocurrency

The advisory pairs the campaign with North Korea's remote IT worker scheme, treating it as the same revenue apparatus seen from the employer's side. Operatives use stolen identity documents and virtual private servers to obscure their locations, working from North Korea, China, Russia, Africa and Southeast Asia, while "laptop farms" in facilitators' homes supply plausible in-country IP addresses. Where placements sour, the agencies say the workers have turned to extortion over payment disputes, website defacement and publication of stolen source code. Japanese investigators dismantled one domestic laptop farm and traced several hundred million yen abroad.

Guidance for practitioners is that run untrusted interview code only in a virtual machine; treat commands containing curl, base64, -enc, mshta or Invoke-WebRequest with suspicion; open unfamiliar projects in Visual Studio Code's Restricted Mode and inspect .vscode/tasks.json first; and after any infection, assume credentials are lost, rotate wallets from a clean device and reinstall the operating system. Employers are told to check applicant IP addresses against claimed locations and to watch for refusal to meet in person, frequent audio and video freezes, or use of face-swapping software.

Also read: North Korean Cyber Fraud Scheme Targets U.S. Firms, DOJ Indicts Five Individuals

The compliance exposure is significant. The U.S. State Department issued an alert to companies on North Korean IT workers in July, and the Justice Department has prosecuted facilitators and seized laptop farms. Firms that unknowingly pay a North Korean contractor risk sanctions liability, which is why the advisory pushes contractual controls and identity verification, not detection alone. Japan's Foreign Ministry said it would "strengthen its efforts with the ally and like-minded countries and the private sector."

What to watch: whether the attribution is followed by sanctions designations, how npm and code-hosting platforms respond to package delivery at this scale, and whether the ACSC issues Australia-specific guidance for contractor and Web3 employers.

Cisco has released a fix for a maximum-severity flaw in its Identity Services Engine (ISE) platform after confirming the bug was already being exploited by attackers. The vulnerability, tracked as CVE-2026-76460, carries a perfect CVSS score of 10.0 and was patched by Cisco on September 16, 2026.  The flaw sits in an API within Cisco Identity Services Engine and stems from inadequate authentication checks on an API endpoint. Because of this weakness, an unauthenticated attacker could send a specially crafted request to that endpoint and slip past ISE's web-based management interface entirely, gaining unauthorized access to the device without needing valid credentials. 

CISA Adds CVE-2026-76460 to Its Exploited Vulnerabilities List 

Given that Cisco confirmed active exploitation of CVE-2026-76460, the Cybersecurity and Infrastructure Security Agency added the flaw to its Known Exploited Vulnerabilities catalog, a designation reserved for bugs already being weaponized in real-world attacks.  Cisco ISE has long served as one of the most widely deployed network access control tools in enterprise environments, helping organizations manage security policies centrally, confirm the identities of users and devices, and govern who and what can reach different parts of a network. That central role is precisely what makes a bypass of its authentication so dangerous, according to security researchers. 

Experts Warn of Ransomware and Lateral Movement Risk 

John Strand, owner of Black Hills Information Security, echoed the urgency, recommending that affected organizations patch immediately and disconnect ISE instances from direct internet exposure. Strand also warned that the bug could plausibly be leveraged in a ransomware attack.   Even so, he noted that relatively few ISE deployments appear to be directly exposed to the open internet, which limits the number of organizations at risk from an opportunistic, internet-facing attack. His larger concern is what happens after attackers are already inside a network: "I'd be more concerned about it as a post-exploitation lateral movement opportunity.   An attacker gets into the environment through something else, discovers a vulnerable ISE internally, and uses that to expand access or compound the damage." 

Signs of Compromise and Recommended Checks 

Cisco's advisory notes that successful exploitation can hand attackers root-level command execution, meaning they could potentially erase or mask evidence of their intrusion. The company is advising administrators to cross-reference network and firewall logs from outside the affected device to spot unusual activity, including unexpected data transfers to external addresses. Suspicious usernames in the access logs are cited as one possible indicator of compromise, and Cisco recommends checking every node in a distributed deployment. 

Affected Products and Fixed Releases 

The vulnerability affects Cisco ISE and ISE Passive Identity Connector regardless of how a device is configured. No workarounds exist, though Cisco suggests using infrastructure access control lists to restrict management traffic as a temporary mitigation. Fixed releases are available for versions 3.1 through 3.5, including 3.1 Patch 12, 3.2 Patch 11, 3.3 Patch 12, 3.4 Patch 7, and 3.5 Patch 4. Version 3.0 has already reached the end of software maintenance, so organizations still running it are advised to migrate to a supported release.  The issue was discovered during a Cisco Technical Assistance Center support case, and the advisory was published in final form on September 16, 2026. 

The Thinkng Bull cyberattack targeted the Seoul-based math academy headquartered in Daechi-dong. The attack has exposed the personal information of students and parents, drawing scrutiny from South Korea's data protection authorities.   The National Math Academy, which runs roughly 80 branches across the country, disclosed the breach to parents on September 17, confirming that an unidentified hacker had infiltrated its exam reservation website. 

Details of the Thinkng Bull Cyberattack 

According to the academy, the intruder informed staff that they were "securing all data," language that led Thinkng Bull to conclude there was a high likelihood that the personal details of every parent and student who had booked exams through the site had been compromised.  The exam reservation platform held sensitive registration data, including students' names, schools, grade levels and dates of birth, along with parents' phone numbers. In response, the National Math Academy took the reservation site offline and filed a report on the breach with the relevant government authorities.   The academy also said it was compiling documentation to request a formal police investigation into the intrusion. 

Academy's Response and Warnings to Parents 

Following the cyberattack on Thinkng Bull, the academy issued a direct warning to parents about the risk of follow-on scams. It noted that the stolen data could be used for "improper purposes" and cautioned that phone calls or text messages from unfamiliar numbers should be treated as potential voice phishing attempts, urging families to "respond calmly" rather than share further details. Parents were also asked not to hand over any personal information or click on unfamiliar links.  Thinkng Bull apologized for the incident, stating it "deeply" regretted the concern caused and expressed hope that no parents or students would suffer actual harm as a result. The academy said the exam reservation website would remain suspended until it completed measures designed to prevent a similar breach from happening again. 

Regulatory Oversight 

South Korea's Personal Information Protection Commission confirmed that Thinkng Bull reported the leak at approximately 9:00 p.m. on September 17. As of the latest updates, the exact number of affected individuals had not yet been determined.   A commission official said the agency was still verifying the facts of the case, including whether any legal violations occurred, and that the timing of any further investigation would depend on the outcome of that review. 

Background on Thinkng Bull 

Thinkng Bull is known as an advanced math program aimed at elementary school students, with a curriculum that allows children in the 4th through 6th grades to complete three years of middle school math material in 13 months.   The academy was founded in 2005 as a single location in Daechi-dong and has since expanded to more than 80 branches nationwide, including one in Gangnam District's Daechi neighborhood. 

What Comes Next? 

The cyberattack on Thinkng Bull adds to a string of data breaches affecting private education providers in South Korea, raising fresh questions about how academies safeguard the personal records of minors collected through online registration systems.  With the Personal Information Protection Commission still reviewing the incident, further details — including the scope of the breach and any regulatory penalties — are expected to emerge as the investigation proceeds.   In the meantime, the National Math Academy has advised affected families to remain alert to unsolicited contact and to avoid sharing additional personal data until the situation is resolved. 

Iranian cyber actors are using a malware family known as HEAVYGRAM, also tracked as CHOSEN BRICK, to target dissidents, journalists, and opposition groups worldwide, according to a recent FBI FLASH report and a joint advisory from the UK National Cyber Security Centre (NCSC), the US FBI, and the Netherlands' General Intelligence and Security Service (AIVD). The FBI assesses that the malware is deployed on behalf of the Government of Iran's Ministry of Intelligence and Security (MOIS) to collect intelligence, conduct data leaks, and cause reputational harm to targets. The joint advisory adds that Iran almost certainly uses cyber activity to support repression of individuals seen as threats to the regime.

Who Is Being Targeted by Iranian Cyber Actors

The victim profile includes Iranian dissidents, journalists opposed to Iran, members of organizations with views countering Government of Iran narratives, and other individuals the Iranian government perceives as threats. The joint advisory notes that CHOSEN BRICK has been used against individuals in the UK, US, and Netherlands since at least 2025, and that personal details of some victims have appeared on pro-Iranian leak sites, increasing risks to their personal safety. In some cases, Iranian intelligence services have plotted to kidnap or conduct lethal operations against individuals abroad who are seen as enemies of the regime.

Social Engineering and Delivery

Iranian cyber actors rely heavily on social engineering to deliver their malware. Actors communicate with targets through platforms such as Telegram, WhatsApp, and Instagram, often posing as individuals known to the target or as technical support from the messaging platform. The joint advisory notes that actors research targets in advance to appear credible, and often initiate contact through a target's work device before shifting to personal devices if the initial delivery attempt fails or risks detection. Victims are persuaded to download files disguised as legitimate applications, including Pictory, RunwayML, Norton Antivirus, Telegram, Adobe Flash Player, and KeePass, or files appearing to be MRI scan results. The FBI notes that some victims downloaded AnyDesk and provided access strings, while others downloaded malware masquerading as a program installer. The malware targets only the Windows operating system.

Persistence and Telegram Command and Control

Once installed, the malware achieves persistence by adding registry keys, most often within the Run key at HKCU\Software\Microsoft\Windows\CurrentVersion\Run, so it runs automatically at user login. It also adds exclusions to Microsoft Defender antivirus to evade detection. The malware then connects to Telegram command-and-control infrastructure, with each victim device linked to a unique Telegram Bot ID to prevent cross-contamination between victims. Through Telegram, the malware can enumerate running processes and system information, capture screen content, enable the microphone to record audio, copy Telegram and WhatsApp data from browsers, download additional files, delete files, steal email content, and wipe the computer system. Files and data are exfiltrated through the Telegram bot and cloud object stores such as VultrObjects and StorjShare. Recent variants use HTTPS/SOCKS5 proxies to obscure the use of Telegram bots.

FBI Recommendations

The FBI recommends caution regarding communications from unknown individuals, keeping devices updated, downloading software only from trusted sources, enabling antivirus software, using strong unique passwords with multi-factor authentication, and reporting suspicious messages. The NCSC, FBI, and AIVD jointly recommend phishing-resistant multi-factor authentication, application allowlisting, email scanning security features, and endpoint and network monitoring. Organizations concerned about a potential HEAVYGRAM or CHOSEN BRICK infection are advised to contact their IT providers and check both corporate and personal devices, since the actors also target individuals outside their workplace systems.

The European Commission has adopted the EU KIDS Act, a proposal that would introduce new age restrictions and safety requirements for online platforms used by children across the European Union. The proposal would prevent social media platforms from allowing children under 13 to access their services and establish an EU-wide minimum age of 15 for minors to open accounts independently.

The EU KIDS Act would also shift the burden of proof to online service providers, requiring them to demonstrate that their services are safe by design and appropriate for younger users.

Commission President Ursula von der Leyen said the proposal is intended to give parents greater control over how children use digital services.

“Today, our children are engaging with the most sophisticated technologies ever created. Technology that was never created with their wellbeing in mind. Our KIDS Act is reversing the burden of proof - it is for platforms to show they are safe by design. And we put parents back in the driving seat, giving them the tools to help their children navigate a safer online world,” von der Leyen said.

EU KIDS Act Sets New Social Media Age Limits

Under the EU KIDS Act, children between 13 and under 15 would not be able to create autonomous social media accounts. Instead, they could use mini accounts managed through a parent or guardian's account, with access limited to age-appropriate social media and video-sharing services.

These mini accounts would include safeguards such as restricted social contacts and screen time of up to one hour per day.

Children between 3 and under 13 would not be permitted to access social media. They could, however, use specially designed child-friendly video-sharing services through guardian-managed accounts. Parents would also receive tools to restrict an adult device to those services when it is given to a child, with exposure limited to a maximum of one hour per day.

Safety-by-Design Rules Target Addictive Features

The proposal would introduce safety-by-design obligations for online services offering social media, video sharing, online video games, AI companions and chatbots to users under 18.

The proposed rules would prohibit addictive features and profiling-based recommendation feeds that can draw minors toward harmful content. Infinite scrolling without stopping points, reward mechanisms and push notifications during sleeping hours would also be restricted.

The proposal would further prohibit unsolicited contact from strangers. AI companions and chatbots would have to be turned off by default and could not simulate interpersonal relationships in ways that create emotional dependency.

Minor profiles would be private by default, while geolocation, camera and microphone access would be turned off. Platforms would also need to provide tools allowing minors to block or mute users, manage their time and control, tune and reset recommender systems.

Privacy-Preserving Age Assurance Proposed

The EU KIDS Act would require online services and app stores to use age assurance tools. The proposal points to the EU age verification app as one possible approach, with the system designed not to retain identity documents or biometric data.

Social media and video-sharing platforms would be expected to perform age verification when users create new accounts. For existing accounts, providers would estimate a user's age using reasonable indicators, such as the account creation date or credit card details.

EU Plans Faster Enforcement

The proposal would require very large online platforms to prove that their services are safe by design. Providers would submit compliance plans to the European Commission and an independent auditor, which would assess new services, features or functionalities.

If the Commission identifies shortcomings based on an auditor's report, it could require providers to propose corrective measures. The enforcement framework would build on existing structures under the Digital Services Act and Artificial Intelligence Act.

For noncompliance, the proposal includes expedited enforcement procedures under which the Commission should conclude investigations within 90 days.

Henna Virkkunen, Executive Vice-President for Tech Sovereignty, Security and Democracy, said the proposal would allow children to use digital services while increasing safeguards and accountability for platforms.

“With our EU KIDS Act, we make sure our children can enjoy their rights online - to learn, stay connected and explore. We also give parents the tools to accompany their kids to navigate a safe digital world. All of this by keeping the pressure high on platforms, because they must prove their services do not harm,” Virkkunen said.

EU KIDS Act Moves to Parliament and Council

The legislative proposal has now been submitted to the European Parliament and the Council for examination and adoption.

The proposal follows recommendations from the European Commission's Special Panel on Child Safety Online, which included more than 60 experts from areas including civil society, education, public health, child psychiatry and technology.

The Commission said public support for stronger online protections is also growing. According to the 2026 Special Eurobarometer on the Digital Decade, 92% of Europeans consider strengthening protection for children and young people online a top policy priority.

Cyble, a global AI-native cybersecurity company, has signed a Memorandum of Understanding (MOU) with the Cyber Security Council of the United Arab Emirates to strengthen national threat intelligence capabilities. The agreement was announced in Abu Dhabi in September 2026 and formalizes a strategic collaboration between Cyble and the federal entity responsible for strengthening the UAE's cybersecurity posture.

The MOU marks a significant milestone in Cyble's engagement with the UAE's national cybersecurity ecosystem, bringing together a global cybersecurity provider and a national authority tasked with protecting the country's digital environment.

Focus on National Threat Intelligence

The MOU establishes a framework for Cyble to support the Council's mission of protecting the UAE's digital infrastructure. The collaboration centers on delivering advanced threat intelligence, early warning capabilities, and actionable insights into emerging cyber risks facing government and critical infrastructure entities across the country. These focus areas are intended to help the Council anticipate and respond to threats before they escalate into significant incidents.

The partnership reflects the UAE's continued commitment to building a resilient, forward-looking cybersecurity posture at the national level. It also positions Cyble as a trusted technology partner in that effort.

By working closely with the Council, Cyble aims to strengthen visibility into threat actors, malicious campaigns, and vulnerabilities that could affect the UAE's most critical systems and services. This visibility is expected to support the Council's broader objective of maintaining a secure and stable digital environment for government and critical infrastructure operators.

Leadership Statements on the MOU

Beenu Arora, CEO and Co-Founder of Cyble, described the agreement as more than a routine partnership announcement. He called it a strategic validation of Cyble's capabilities at the national cybersecurity level and a potential gateway to the UAE government and critical infrastructure ecosystem. This is not just a partnership announcement. It is a strategic validation of Cyble's capabilities at the national cybersecurity level and potentially a gateway to the UAE government and critical-infrastructure ecosystem," said Arora.

Arora added that working alongside the Cyber Security Council underscores the trust placed in Cyble's threat intelligence platform and reflects the growing importance of proactive, intelligence led defense in protecting nations against increasingly sophisticated cyber threats.

Dr. Mohamed Al Kuwaiti, Chairman of the Cyber Security Council of the United Arab Emirates, said the Council is committed to building strategic partnerships with world class technology providers that strengthen the UAE's cyber resilience at every level. He noted that the MOU with Cyble reflects a shared commitment to safeguarding the nation's digital infrastructure through advanced threat intelligence and proactive, intelligence led defense, ensuring the UAE remains at the forefront of global cybersecurity readiness.

"This MOU with Cyble reflects our shared commitment to safeguarding the nation's digital infrastructure through advanced threat intelligence and proactive, intelligence-led defense — ensuring the UAE remains at the forefront of global cybersecurity readiness," said Dr. Mohamed Al Kuwaiti.

The agreement builds on Cyble's broader strategy of deepening engagement with government and regulatory bodies across the Middle East. It reinforces the company's position as a trusted partner to national cybersecurity authorities navigating an increasingly complex threat landscape, and it adds to a growing list of engagements between Cyble and public sector organizations in the region.

Role of Cyble Vision and Blaze AI

Cyble Vision, the company's threat intelligence platform powered by Blaze AI, will play a central role in the collaboration. The platform is designed to deliver rapid detection, automated analysis, and contextual insights into ransomware, phishing, fraud, and other evolving threats relevant to national security priorities. Through this platform, Cyble intends to provide the Council with tools that support faster identification and analysis of threats targeting the UAE's digital infrastructure.

Cyble is a global, AI-native cybersecurity company specializing in threat intelligence, digital risk protection, and dark web monitoring. Its flagship platform, Cyble Vision, powered by Blaze AI, delivers rapid threat detection, automated analysis, and contextual insights into ransomware, phishing, and fraud.

Media Contacts: enquiries@cyble.com | +1 678 379 3241 | www.cyble.com

A newly documented security flaw in Parallels Desktop, identified as CVE-2026-90894 and nicknamed "ParaShells," could let any local account on a Mac escalate to full root control of the host machine, according to researchers at JFrog.  The risk is greatest on developer machines, where something as ordinary as a compromised Homebrew formula or a malicious npm preinstall script could serve as the entry point for escalating from a standard user to complete system control.   Shared computers with multiple local accounts — think university labs or office workstations — face similar exposure, said Yuval Moravchick, who leads JFrog's vulnerability research team.  Once an attacker reaches root, Moravchick said, they can swap out system software, access other users' files, and maintain persistence through launchd. 

Inside the ParaShells Exploit Chain 

CVE-2026-90894 is an argument-injection bug confirmed to affect Parallels Desktop for Mac version 26.4.0 running on Apple silicon. Earlier releases haven't been ruled out, but JFrog's team didn't test them.  Three conditions have to line up for ParaShells to work: the vulnerable version of Parallels needs to be installed, though no virtual machine actually has to be running; the prl_disp_service — which launches automatically as root via its daemon — must be active, along with its accompanying socket at /var/run/prl_disp_service.socket; and the attacker needs the ability to execute code under any low-privileged local account.  Moravchick described the mechanics as a combination of weaknesses stacking together: a Unix socket left world-writable, a client-authentication scheme that verifies peer credentials instead of checking a Team ID, and an argument-injection flaw in how the software unpacks appliance files. "The chain is short," he said. "A world-writable Unix socket, a login that trusts peer credentials rather than a Team ID, and an appliance unpack path that builds tar arguments using Qt string splitting." A stray quote character in the parent path allows the attacker to inject the flag "–use-compress-program=", causing macOS's tar utility to execute the attacker's script with root privileges. 

Disclosure and Fix for CVE-2026-90894 

JFrog reported the ParaShells vulnerability to Parallels' parent company, Alludo, which patched the issue in Parallels Desktop version 27.0.0, released in early September 2026. The company published its technical write-up two weeks after the patch shipped, but withheld the one-line proof-of-concept script it had developed to demonstrate the exploit.  Moravchick recommended that organizations identify every Mac running Parallels Desktop, restrict local account permissions on those systems, and update to version 27.0.0 or later to close off CVE-2026-90894.