OpenAI warns that frontier AI models could escalate cyber threats, including zero-day exploits. Defense-in-depth, monitoring, and AI security by design are now essential.
The post As Capabilities Advance Quickly OpenAI Warns of High Cybersecurity Risk of Future AI Models appeared first on Security Boulevard.
I have long maintained that smart contracts are a dumb idea: that a human process is actually a security feature.
Here’s some interesting research on training AIs to automatically exploit smart contracts:
AI models are increasingly good at cyber tasks, as we’ve written about before. But what is the economic impact of these capabilities? In a recent MATS and Anthropic Fellows project, our scholars investigated this question by evaluating AI agents’ ability to exploit smart contracts on Smart CONtracts Exploitation benchmark (SCONE-bench)a new benchmark they built comprising 405 contracts that were actually exploited between 2020 and 2025. On contracts exploited after the latest knowledge cutoffs (June 2025 for Opus 4.5 and March 2025 for other models), Claude Opus 4.5, Claude Sonnet 4.5, and GPT-5 developed exploits collectively worth $4.6 million, establishing a concrete lower bound for the economic harm these capabilities could enable. Going beyond retrospective analysis, we evaluated both Sonnet 4.5 and GPT-5 in simulation against 2,849 recently deployed contracts without any known vulnerabilities. Both agents uncovered two novel zero-day vulnerabilities and produced exploits worth $3,694, with GPT-5 doing so at an API cost of $3,476. This demonstrates as a proof-of-concept that profitable, real-world autonomous exploitation is technically feasible, a finding that underscores the need for proactive adoption of AI for defense...
The post AIs Exploiting Smart Contracts appeared first on Security Boulevard.
I have long maintained that smart contracts are a dumb idea: that a human process is actually a security feature.
Here’s some interesting research on training AIs to automatically exploit smart contracts:
AI models are increasingly good at cyber tasks, as we’ve written about before. But what is the economic impact of these capabilities? In a recent MATS and Anthropic Fellows project, our scholars investigated this question by evaluating AI agents’ ability to exploit smart contracts on Smart CONtracts Exploitation benchmark (SCONE-bench)a new benchmark they built comprising 405 contracts that were actually exploited between 2020 and 2025. On contracts exploited after the latest knowledge cutoffs (June 2025 for Opus 4.5 and March 2025 for other models), Claude Opus 4.5, Claude Sonnet 4.5, and GPT-5 developed exploits collectively worth $4.6 million, establishing a concrete lower bound for the economic harm these capabilities could enable. Going beyond retrospective analysis, we evaluated both Sonnet 4.5 and GPT-5 in simulation against 2,849 recently deployed contracts without any known vulnerabilities. Both agents uncovered two novel zero-day vulnerabilities and produced exploits worth $3,694, with GPT-5 doing so at an API cost of $3,476. This demonstrates as a proof-of-concept that profitable, real-world autonomous exploitation is technically feasible, a finding that underscores the need for proactive adoption of AI for defense.
Security defenders are girding themselves in response to the disclosure of a maximum-severity vulnerability disclosed Wednesday in React Server, an open-source package that’s widely used by websites and in cloud environments.
The vulnerability is easy to exploit and allows hackers to execute malicious code on servers that run it. Exploit code is now publicly available.
React is embedded into web apps running on servers so that remote devices render JavaScript and content more quickly and with fewer resources required. React is used by an estimated 6 percent of all websites and 39 percent of cloud environments. When end users reload a page, React allows servers to re-render only parts that have changed, a feature that drastically speeds up performance and lowers the computing resources required by the server.
© Getty Images
A sprawling infrastructure that has been bilking unsuspecting people through fraudulent gambling websites for 14 years is likely a dual operation run by a nation-state-sponsored group that is targeting government and private-industry organizations in the US and Europe, researchers said Wednesday.
Researchers have previously tracked smaller pieces of the enormous infrastructure. Last month, security firm Sucuri reported that the operation seeks out and compromises poorly configured websites running the WordPress CMS. Imperva in January said the attackers also scan for and exploit web apps built with the PHP programming language that have existing webshells or vulnerabilities. Once the weaknesses are exploited, the attackers install a GSocket, a backdoor that the attackers use to compromise servers and host gambling web content on them.
All of the gambling sites target Indonesian-speaking visitors. Because Indonesian law prohibits gambling, many people in that country are drawn to illicit services. Most of the 236,433 attacker-owned domains hosting the gambling sites are hosted on Cloudflare. Most of the 1,481 hijacked subdomains were hosted on Amazon Web Services, Azure, and GitHub.
© Getty Images
Apple is now offering a $2M bounty for a zero-click exploit. According to the Apple website:
Today we’re announcing the next major chapter for Apple Security Bounty, featuring the industry’s highest rewards, expanded research categories, and a flag system for researchers to objectively demonstrate vulnerabilities and obtain accelerated awards.
- We’re doubling our top award to $2 million for exploit chains that can achieve similar goals as sophisticated mercenary spyware attacks. This is an unprecedented amount in the industry and the largest payout offered by any bounty program we’re aware of and our bonus system, providing additional rewards for Lockdown Mode bypasses and vulnerabilities discovered in beta software, can more than double this reward, with a maximum payout in excess of $5 million. We’re also doubling or significantly increasing rewards in many other categories to encourage more intensive research. This includes $100,000 for a complete Gatekeeper bypass, and $1 million for broad unauthorized iCloud access, as no successful exploit has been demonstrated to date in either category.
- Our bounty categories are expanding to cover even more attack surfaces. Notably, we’re rewarding one-click WebKit sandbox escapes with up to $300,000, and wireless proximity exploits over any radio with up to $1 million.
- We’re introducing Target Flags, a new way for researchers to objectively demonstrate exploitability for some of our top bounty categories, including remote code execution and Transparency, Consent, and Control (TCC) bypasses and to help determine eligibility for a specific award. Researchers who submit reports with Target Flags will qualify for accelerated awards, which are processed immediately after the research is received and verified, even before a fix becomes available.
Login