OpenAI matched Anthropic's promise to put outside evaluators inside the company
Sam Altman said OpenAI will give independent evaluators employee-like access to its systems. It matches the commitment Dario Amodei made for Anthropic in his Saturday essay. Writing on X, Altman said he agreed that the industry needs to pace the frontier. He said the question had been a main topic inside OpenAI for weeks. He called committing to independent evaluators with employee-like access a good idea, and said the company would share more later.
Verified 7:20 AM PDT · 4 original sources
Elon Musk replied to the essay with three words, writing that Amodei is right. Altman also clarified that pacing does not mean stopping, and that no amount of American competitive pressure justifies recklessness.
As of September 14, OpenAI has not published the scope of the access, the evaluators it would admit or a timetable. The company ruled out a public offering in 2026 last week.
These are company commitments with no contract attached. No evaluator has been named, and no agreement has been published. Nothing states whether an outside team could release a finding the company dislikes. Nor does anything say whether it would see training runs before a release decision. A commitment made in a social post is also easy to soften later without a written breach.
Watch for a joint statement from the frontier labs that names participants and defines what access includes. The measurable test is whether an evaluator can publish without company approval and whether any lab discloses a release it delayed after that review.
Audit the story
Original sources
Company claims remain company claims. Follow the reporting and judge the evidence directly.
- Sam AltmanPost committing OpenAI to independent evaluators with employee-like access ↗
- International Business Times UKAltman Matches Anthropic's AI Auditor Pledge While Musk Offers Three-Word Slowdown Backing ↗
- MedianamaAnthropic wants independent evaluators inside frontier AI labs; OpenAI agrees ↗
- ThePrintAltman and Musk back Anthropic CEO's call to slow AI race ↗
Continue the edition