OpenAI has formed a new Safety and Security Committee less than two weeks after dissolving its previous "Superalignment Team" tasked with protecting humanity from AI risks. The new committee is led by OpenAI board members including the board chair, two other board members, and CEO Sam Altman, raising concerns about self-policing. It follows the high-profile resignations of two members of the previous safety team, who expressed concerns that safety was taking "a backseat" to products at OpenAI. The new committee plans to evaluate OpenAI's safety processes over 90 days and make recommendations, but is led by the same leadership developing new models.
