Anthropic's Dario Amodei calls to slow AI, opens models to third-party evaluators like METR
The Verge AI · rss · 2026-09-13
Anthropic CEO Dario Amodei published a wide-ranging essay arguing it's time to slow AI development, proposing a three-step plan to "pace the frontier" — slowing training and development to give companies time to build safeguards and regulators time to evaluate models. Step one is underway unilaterally: Anthropic will give third-party evaluators like METR wide-ranging access to its models to verify adherence to safety practices and commitments. Step two would involve industry-wide coordination, likely with governments.
More from Companies & People
- Anthropic Alignment Lead Amanda Askell Sparks Firestorm Over Remarks on Discrimination — SumitGup · 2026-09-14
- Hackathon team builds hawker digitalization B2B prototype in a day with OpenAI Astra — gabrielchua · 2026-09-14
- 2022 LessWrong quote resurfaces: implementing a transformer in PyTorch once got you recruited by Redwood — panickssery · 2026-09-14
- Alexandr Wang: Meta spent a year secretly tuning its models for the Muse personal agent — alexandr_wang · 2026-09-14
- Scale CEO Alexandr Wang backs Meta Muse as shifting economic surplus to consumers — alexandr_wang · 2026-09-14
- Anthropic reportedly signed a $13.7B, 6-year compute deal with RUM Group's Georgia site — rohanpaul_ai · 2026-09-14