Google DeepMind CEO Demis Hassabis has proposed a US-led regulatory body modeled after FINRA to conduct safety screenings on frontier AI systems. The plan aims to establish a systematic framework for managing risks, including potential cyber and biological threats.

Fidji Simo, OpenAI's head of AGI deployment, is transitioning from her full-time role to an advisory position to focus on her health. Her responsibilities will be redistributed among the existing leadership team.