OpenAI Announces Safety and Security Committee Amid New AI Model Development
28 May 2024 at 11:09
OpenAI announced a new safety and security committee as it begins training a new AI model intended to replace the GPT-4 system that currently powers its ChatGPT chatbot.
The San Francisco-based startup announced the formation of the committee in a blog post on Tuesday, highlighting its role in advising the board on crucial safety and security decisions related to OpenAI’s projects and operations.
The creation of the committee comes amid ongoing debates about AI safety at OpenAI. The company faced scrutiny after Jan Leike, a researcher, resigned, criticizing OpenAI for prioritizing product development over safety. Following this, co-founder and chief scientist Ilya Sutskever also resigned, leading to the disbandment of the "superalignment" team that he and Leike co-led, which was focused on addressing AI risks.
Despite these controversies, OpenAI emphasized that its AI models are industry leaders in both capability and safety. The company expressed openness to robust debate during this critical period.
OpenAI's Safety and Security Committee Composition and Responsibilities
The safety committee comprises company insiders, including OpenAI CEO Sam Altman, Chairman Bret Taylor, and four OpenAI technical and policy experts. It also features board members Adam D’Angelo, CEO of Quora, and Nicole Seligman, a former general counsel for Sony."A first task of the Safety and Security Committee will be to evaluate and further develop OpenAI’s processes and safeguards over the next 90 days."The committee's initial task is to evaluate and further develop OpenAI’s existing processes and safeguards. They are expected to make recommendations to the board within 90 days. OpenAI has committed to publicly releasing the recommendations it adopts in a manner that aligns with safety and security considerations. The establishment of the safety and security committee is a significant step by OpenAI to address concerns about AI safety and maintain its leadership in AI innovation. By integrating a diverse group of experts and stakeholders into the decision-making process, OpenAI aims to ensure that safety and security remain paramount as it continues to develop cutting-edge AI technologies.