Anthropic and OpenAI sought control over outside AI safety reviews
Two AI firms reportedly wanted significant say over what independent evaluators could publish, over concerns that findings might expose trade secrets.
By Standard Topic. AI-generated from the reporting listed under Coverage. Published . Updated .
This story is no longer being updated. It reflects reporting up to .
What happened
Anthropic and OpenAI have reportedly sought significant control over what outside evaluators can publish, according to people familiar with the discussions. The companies were concerned that findings could reveal trade secrets.
Why it matters
The terms under negotiation could shape what the public learns about AI risks. Researchers need access to unreleased systems, but company approval and redaction rules can limit what they tell the public.
Timeline
- Oct 5: Anthropic and OpenAI privately sought strict editorial authority over what outside reviewers could publish.
- Oct 13, 6:07 AM ET: Reuters reported that the two companies sought significant control over what outside evaluators could publish, citing concerns that findings could reveal trade secrets.
- Oct 14, 12:45 AM ET: The reported requests were described as contradicting September pledges from both CEOs that evaluators would work without company approval.
Coverage
- An update on independent evaluators (Anthropic)
- Anthropic and OpenAI Are Negotiating Publication Rights of External AI Safety Reviews (winbuzzer.com)
- Anthropic and OpenAI Pushed for Veto Rights Over Safety Reports While Pledging Independence (techtimes.com)
- OpenAI wants independent testing. Who controls the findings? (thetechpencil.com)
- AI Labs Having Second Thoughts on Independent Safety Evaluations (PYMNTS)
- Anthropic and OpenAI have reportedly sought significant control over what outside evaluators can publish,… (Newsquawk)
- Exclusive-Behind AI industry's call for independent reviews, a fight over what gets made public (Reuters)
- Exclusive-Behind AI industry’s call for independent reviews, a fight over what gets made public (Reuters)
More from artificial-intelligence
- Anthropic co-founder urges binding rules for advanced AI
- China proposes export controls on dozens of technologies
- Accenture caps employee spending on Anthropic's Claude at $29 monthly
- Anthropic's IPO outlook cools as investors weigh lower valuation
- G7 finance chiefs agree to AI cyberattack drills
- OpenAI research model takes funds from unowned crypto wallet
All artificial-intelligence news · artificial-intelligence story archive