Watchtower
Watchtower tracks when AI companies change their safety policies — and when they break them.
Released an acceptable usage policy, along with a reporting email for security vulnerabilities.
Cognition
Sep 4, 2024
On September 4, 2024, Cognition AI released a new acceptable use policy, along with a reporting email for security vulnerabilities.
This release came in the wake of The Midas Project's campaign against Cognition for (previously) having never discussed safety or appropriate usage.
However, it was also released in a largely unpolished format with multiple errors and typos, including misspelling the name of their main product, Devin. These errors have since been fixed.
OpenAI
Aug 18, 2026
OpenAI updated its Model Spec, the document outlining intended model behavior
xAI
Aug 17, 2026
xAI updated Grok 4.6's model card after release. A changelog was included.
Aug 14, 2026
Google updated its Gemini 3.7 Flash model card after publication, making changes to language in the “Key Results for Gemini 3.7 Flash” column in the Frontier Safety Assessment section
OpenAI
Aug 3, 2026
OpenAI updated its system card for two of its models: GPT 5.6 and GPT Live.
xAI
Jul 20, 2026
xAI made changes throughout the model card for Grok 4.5, including some that appear to be persistent errors
Frequently asked questions
Have more questions? Our team is happy to help, contact us.