Anthropic and OpenAI want to embed safety evaluators. Will they really be independent?
TechCrunch AI · rss · 2026-09-17
Anthropic and OpenAI are both moving to embed independent safety evaluators inside their AI labs, granting outside researchers unprecedented access to model evaluations. Researchers welcome the access but caution that meaningful oversight requires transparency, genuine independence, and eventually regulation—voluntary internal arrangements alone won't suffice.
More from Companies & People
- Databricks rolls out GPT-6 Astra to 3,500 engineers, coding spend jumps 60% — nickbaumann_ · 2026-09-17
- Runway hires HCI researcher nithinaray to work on AI video and world models — c_valenzuelab · 2026-09-17
- Pick a lab whose people you like — agents now do the work while you hang out — joannejang · 2026-09-17
- OpenAI hints at millennium problem breakthrough ahead of Sept 29 DevDay — haider1 · 2026-09-17
- Former Waymo CFO Elisa de Martel joins self-driving startup Wayve — alexgkendall · 2026-09-17
- Sentry CEO plans full-time hires for Junior, its AI agent project — zeeg · 2026-09-17