22 countries call for frontier AI controls, still pushing standards and evals
DavidSKrueger · x · 2026-09-22
22 countries, including Germany, the Netherlands, and the UAE, have jointly called for controls on frontier AI. David Krueger notes they are still asking mainly for standards and evals, which he argues is insufficient—hoping they can be educated to go further.
More from Safety
- AI Fear Report tracks 1,815 AI bills, finds deepfake fear tops the regulatory index — beffjezos · 2026-09-22
- 670,000 agent skills, no trust layer: bot scan finds 69% never reliably fire — markjeffrey · 2026-09-22
- Insider: Frontier Model Forum is herding cats, lacks consensus to publish anything — evijit · 2026-09-22
- OpenAI's AGI plan: build an automated AI researcher, then iterate on alignment with it — trevposts · 2026-09-22
- US government lists first dedicated voice AI vendor under FedRAMP 20x — Prestigious_Bet2215 · 2026-09-22
- Doe v. GitHub AI case sees a meaningful ruling on Copilot output claims — technollama · 2026-09-22