AI-Generative Signal
The daily Signal newsletter, archived. Newest edition first. · The Move archive
← Back to the latest edition
September 13, 2026Anthropic’s Dario Amodei Proposes ‘Pacing’ Frontier AI Development With Embedded Outside Review
Anthropic has proposed a three-stage framework for pacing frontier AI development. Its immediate commitment is to invite embedded third-party evaluators; shared industry limits and international coordination remain proposals that would require broader adoption.
September 12, 2026New Mexico Supreme Court Fines Lawyer $5,000 Over AI-Fabricated Witnesses in Murder Appeal
New Mexico’s highest court fined an attorney $5,000 and referred him for disciplinary investigation after an AI-assisted murder-appeal brief reportedly included fabricated witnesses and false testimony. The case clarifies a core governance principle for regulated workflows: generative AI can assist…
September 11, 2026OpenAI’s Navier–Stokes Claim Is a Test of AI Math—and of Scientific Trust
OpenAI’s Navier–Stokes announcement is a potential milestone for agentic AI research, but not a settled theorem. The public Lean proof makes the work more inspectable, while the related dispute over unpublished research inputs exposes a governance problem that formal verification…