METR Reviews Anthropic's Automated R&D Risk Claims With Access to Nonpublic Report
DKokotajlo · x · 2026-11-06
METR published an independent review of the "Risks from automated R&D" section of Anthropic's August 2026 Risk Report, working from the nonpublic version of the document.
The review focuses on the report's claims about catastrophic risks from automated AI R&D and on whether Anthropic has crossed relevant capability thresholds. A rare third-party audit of a frontier lab's safety commitments.
More from AGI Musings
- Personal Websites Will Be the New Facebook Profile in the Agent Era — signulll · 2026-11-06
- Hedge drift in AI progress debates: METR's caveats vanish as AECI scores hit 1.9x — sebkrier · 2026-11-06
- O'Reilly: Software's Core Artifact Is Shifting From Code to Prompts — luisdans · 2026-11-06
- RSI isn't about acceleration—it's removing humans as the bottleneck — menhguin · 2026-11-06
- Stop Measuring AI in Gigawatts and GPUs, Architect David Linthicum Argues — DavidLinthicum · 2026-11-06
- Forecast model: 2x progress rate in Dec 2026 implies ~4.5x capability uplift — arthurcolle · 2026-11-06