Follow-up: cross-lab eval ban framed as "data hygiene," not eval secrecy

anpaure · x · 2026-11-16

Responding to the report that OpenAI can't run safety evals on Anthropic models, anpaure frames it as "data hygiene" — not running anyone else's models at work — rather than hiding the eval itself, suggesting third-party auditors should ideally handle cross-lab evaluations.

Related event: OpenAI apparently barred from running safety evals on Anthropic models, sparking cross-lab debate(6 posts)→

Original post →

More from Safety

Safety channel →