What else happened on Oct 22nd
The October 22 posts converge on one question: who checks the AI labs, and how far that checking can be trusted. Participants in the Anthropic Fellows program attacked Claude Code's blocking monitor and their own reconstruction of the Guardian monitor from OpenAI's Codex; Fathom proposes that independent auditors get a license from the state; and the French center CeSIA looks at what leverage Britain actually has over the labs.