Good morning! Today's big story is a new top model. Anthropic released Claude Fable 5.2 and Claude Mythos 5.2, and Fable now leads the Artificial Analysis ranking. It lands a few days after METR said Anthropic's models improve about twice as fast as a year ago. It also lands in the week Anthropic is expected to set its IPO price range.

Look at the rest of the news and one word keeps coming back: trust. Nvidia's board is investigating its CEO over ties to a suspected Chinese spy, and the warning signs were there a year ago. Coding agents keep touching our keys and passwords. Gemini faces a new claim about human reviewers, and users ask if Opus 5.5 was quietly changed. Each time, the public learns late, and the only ones who can really check are the companies themselves.

That is my worry with this speed: the models change every few weeks, the rules and the checks move much slower. Money from a stock market sale will not slow anything down. Australia's push for incident reporting with a deadline is the kind of answer we need more of. To watch: Anthropic's valuation range, Nvidia's board, and whether Google and Anthropic answer this week's claims.

Today's Morning Tokens:

1. Anthropic launches Claude Fable 5.2 and Mythos 5.2, days before setting its IPO price
2. Nvidia scandal: Huang admits affair, stock slides, lawyers circle
3. Your coding agents and your keys: Claude Code secrets hit the news, ChatGPT Desktop questioned
4. New claim: Gemini sends personal Drive files and emails to human reviewers
5. Was Opus 5.5 quietly weakened on November 3? Users ask, nobody can check
6. Australia moves toward deadlines for reporting AI incidents
7. Meta's Muse Ember: a cheap small model for the work agents repeat all day

Read it all, with the sources: https://aidailygrind.com/2026-11-10.html

Read by an AI voice. Stories are picked and written with the help of AI tools, from public news sources. Always check the sources.