OpenAI Faces Turmoil as Safety Team Disbands Amid Resignations of Sutskever and Leike

May 17, 2024, 05:10 PM

OpenAI has recently experienced significant internal turmoil with the resignation of key figures from its safety-focused Superalignment team. Notably, co-founder Ilya Sutskever and head of alignment Jan Leike have both departed. The Superalignment team, which was responsible for mitigating the existential dangers of AI, has been disbanded, with its members either leaving the company or being reassigned to other research groups. Leike cited a shift in OpenAI's priorities, stating that the company has increasingly focused on 'shiny products' over AI safety. This shift has led to resource constraints for the safety team, which struggled to secure the promised 20% of the company's compute resources. The disbandment of the Superalignment team raises concerns about OpenAI's commitment to AI safety and its long-term vision for managing the risks associated with advanced AI development. WIRED confirmed the disbandment, and concerns have been raised about CEO Sam Altman's leadership.