Watchtower
Watchtower tracks when AI companies change their safety policies — and when they break them.
Updated its preparedness framework
OpenAI
Apr 15, 2025
OpenAI updated its preparedness framework, its first major update since the policy was originally released in December 2023.
Analysis of the update, comparing the new and old policies, will be added to the page soon.
A document of the new policy (nearly a complete rewrite) can be found at this link.
Meta
Oct 23, 2026
Meta quietly updated Wonder’s model card nearly two weeks after the model’s release, with still no sign of a safety and preparedness report for the model.
OpenAI
Oct 21, 2026
OpenAI updated its Usage policies. The new policies will go into effect on Nov. 20.
Oct 13, 2026
Google launched Gemini 4 Argon along with the first “full” safety and preparedness report made under its Frontier Safety Framework (v3.1), published in April 2026.
Meta
Oct 12, 2026
Meta publicly launched Muse Wonder with a 13-page model card and without concurrently publishing a safety and preparedness report.
Anthropic
Oct 8, 2026
Anthropic updated its Usage Policy. The new policy will go into effect on Nov. 12.
OpenAI
Oct 7, 2026
OpenAI published a new system card for GPT-6 Sol and GPT-6 Luna, covering new versions of the two models available in the basic ChatGPT interface.
Frequently asked questions
Have more questions? Our team is happy to help, contact us.