OpenAI Security Personnel Departing from the Organization
OpenAI’s Superalignment Team Leadership Resignation
OpenAI established its Superalignment team nearly a year ago with the primary objective of managing hypothetical superintelligent AI systems to prevent any potential threats to humans. The recent resignation of the team’s leaders has sparked concerns among many individuals, questioning the necessity of such a team’s existence.
The departure of the Superalignment leadership is part of a series of significant exits from the company, some originating from Sutskever and Leike’s security-focused team. In November 2023, Sutskever and the OpenAI board spearheaded an unsuccessful attempt to remove CEO Sam Altman. Subsequently, several OpenAI staff members departed, engaging in discussions on AI safety or contributing to essential security teams.
Despite Sutskever’s subsequent apology and joint letter signing with 738 OpenAI employees in support of Altman and president Greg Brockman, some former staff members, including Jan Leike, Leopold Aschenbrenner, and William Saunders from the Superalignment team, did not sign the letter or took time to do so. Notable AI researcher Andrej Karpathy, along with former OpenAI employees Daniel Kokotajlo and Cullen O’Keefe, also left the company and did not appear in the letter supporting the leadership.
Gizmodo has attempted to reach out to OpenAI for comments on the future leadership of the Superalignment team but has not received a response. The issue of security at OpenAI has historically been contentious, prompting Dario and Daniela Amodei to depart in 2021 and establish Anthropic, along with nine other former OpenAI staff members. Security concerns also played a role in the removal of Altman and Brockman by non-profit OpenAI’s board members, who were subsequently replaced by well-known Silicon Valley entrepreneurs.
Despite the departures, OpenAI continues to have a dedicated team working on security, aligning with the company’s mission to develop AGI safely for the betterment of humanity. Gizmodo has compiled an updated list of notable AI safety advocates who have left OpenAI following Altman’s removal.
