OpenAI's GPT-6 Astra is the first model the company has rated at the Critical level for cybersecurity under its Preparedness Framework. The public version is limited to reviewing code for security problems and helping patch them. OpenAI says it refuses requests to build proof-of-concept exploits.
Nico Park · Monday 14 September 2026
Sources
This desk has not corrected this file. Corrections.






The paper
Comments
No notes on this story yet.
Sign in to comment