Cohere CEO Aidan Gomez: third-party AI audits are power projection, not real fixes
yacineMTB · x · 2026-09-13
Cohere founder Aidan Gomez pushed back on proposals for third-party AI audits, arguing the actual risks are a) OpenAI/Anthropic models hacking when given high-risk tasks in weakly secured sandboxes, and b) Chinese labs open-sourcing increasingly capable models. Neither, he says, is fixed by auditors: the first just requires better sandboxes, the second is a geopolitics problem. He calls third-party audits a means of power projection that arbitrarily raises barriers for lower-resourced players while solving nothing.
Related event: Anthropic CEO Dario Amodei Calls for Slowing Frontier AI Development(203 posts)→
More from Safety
- VLC's VideoLAN blisteringly rebukes French lawmaker over digital tax criticism — ngxson · 2026-10-31
- ChatGPT Daybreak Blue sessions may train OpenAI models even with training opt-out off — littmath · 2026-10-31
- Ramez Naam: cybersecurity needs frontier models used aggressively, not guardrails — Jsevillamol · 2026-10-31
- OpenAI reportedly to train on users' Daybreak sessions with no opt-out — dyn___ · 2026-10-31
- Forethought: US has 9x China's AI compute, can afford to pace the frontier — willmacaskill · 2026-10-31
- unpromptedconf takeaways: security programs face forced change as reachable vulnerabilities can't be risk-accepted — scottleibrand · 2026-10-31