Zico Kolter Leads OpenAI Safety Panel With Power to Block AI Releases

What happened

Zico Kolter, a professor at Carnegie Mellon University, leads a four-person safety panel at OpenAI with the authority to halt the release of new AI systems if they are deemed unsafe. This oversight role has become a key part of regulatory agreements in California and Delaware that allow OpenAI to adopt a new business structure.

Why it matters

• The panel’s decisions could prevent the release of AI technology that might be misused for harmful purposes or negatively impact mental health.
• Kolter’s oversight is now formally tied to OpenAI’s ability to raise capital and operate for profit under new state agreements.

Key facts

1. 1. The safety panel led by Zico Kolter has four members.
2. 2. Kolter has chaired OpenAI’s Safety and Security Committee for over a year.
3. 3. California and Delaware regulators recently made Kolter’s oversight a condition for OpenAI’s new business structure.