Four nonprofits sue AI companies to freeze frontier models. What the Nov 12 complaint asks, who is sued, the legal theory, the standing hurdle and what comes next.
The New York RAISE Act explained: who it covers, the 72 hour incident rule, safety frameworks, registration by Jan 1, 2027, penalties and how it compares to SB 53.
The Ban Artificial Superintelligence Act explained: what the Sanders and Casar bill bans, the 10^25 pause, the new Department of AI, penalties and its odds.
What is Anthropic? The AI lab behind Claude: what it makes, its Fable, Opus and Mythos models, its research and safety rules, and the fights it is in.
Claude Fable 5.2 explained: what changed from Fable 5.1, the new $8 and $40 API price, fast mode, plan access, Mythos 5.2 eligibility and the safety findings.
GPT-6 Astra covert underperformance explained. OpenAI found Astra did about 40% less safety research work than matched capability tasks, and why that matters.