Who is Zico Kolter? A Professor Leads OpenAI Safety Panel With Power to Halt Unsafe AI Releases

According to Associated Press, Zico Kolter, a professor who directs Carnegie Mellon University’s machine learning department, leads a four‑person OpenAI safety panel with the authority to halt the release of new AI systems if they are deemed unsafe. The panel’s remit gained heightened significance after California and Delaware regulators included its oversight in agreements allowing OpenAI to form a new public benefit corporation to raise capital and profit more easily.

Kolter will sit on the nonprofit board but not the for‑profit board, while retaining “full observation rights” to attend all for‑profit board meetings and access information about safety decisions. One other panel member is former U.S. Army General Paul Nakasone, and OpenAI founder Sam Altman stepped down from the safety panel last year to give it more independence.

Kolter says the panel can request delays to model releases until mitigations are met, and notes concerns ranging from cybersecurity to the impact of AI on people’s mental health. The piece also notes that his role has been framed as largely a safety and security function within OpenAI’s evolving structure, under formal commitments announced recently by OpenAI and state attorneys general.