Ryan Greenblatt warns open-weight restrictions would hurt safety research

RyanGreenblatt · x · 2026-07-23

A public back-and-forth centers on whether the administration is preparing heavy restrictions on open-weight AI models.

Related event: GPT-OSS Open-Source and Safety Debate: Risk, Censorship, and Regulation(20 posts)→

Original post →

More from Safety

Safety channel →