Researchers push back: frontier AI labs shouldn't pick their own auditors or questions
agstrait · x · 2026-09-17
Researcher Angela Zhou argues that frontier AI labs cannot be allowed to choose their own auditors — or dictate the questions those auditors ask. The post is a call for more independent external oversight of frontier-lab safety claims, linking to a longer piece, though the tweet itself doesn't elaborate on specifics.
More from Safety
- UK Think Tank Warns a Small Club of US and Chinese Labs Is Stitching Up an AI Super League — alexvoica · 2026-09-17
- 28 fake dating apps, 4,700 AI girlfriends, 25,000 guys paying per message — diegoposts · 2026-09-17
- Cisco's 3B Model Matches GPT-5.5 on Bug Localization (0.223) With Far Fewer False Positives — shashib · 2026-09-17
- Goodfire partners with Baseten and Base Labs to build safety infrastructure for open-source models — Thom_Wolf · 2026-09-17
- Pentagon listed Alibaba as military company in June — yet a federal site tests Qwen3 — 96Stats · 2026-09-17
- OpenAI caught Astra jailbreaking itself via malicious summary instructions, 27 cases found — RexDouglass · 2026-09-17