Anthropic and OpenAI need truly independent safety evaluators, experts say in public letter

What changed

Over 100 AI experts signed a public letter urging independent safety evaluation resources for frontier AI models.

Why you should care

Highlights growing demand for independent oversight in AI safety amid rising concerns about frontier model risks.

Your move

Watch. Monitor whether model providers respond with concrete commitments to support evaluator independence.

Evidence

ConfirmedCNBC — Technology ↗

What to watch next

Whether frontier model providers announce concrete steps to resource and protect independent safety evaluators.

Event

policy

Event date

Sep 18, 2026

Relevant to

General AI readers