An independent reviewer should be able to say things Anthropic hates

Giving an evaluator deep access is a good step. Letting them publish uncomfortable findings is the real test. Anthropic's October evaluator update says METR can publish without an Anthropic veto. It also says Anthropic can require redactions in predefined categories and ultimately decides what falls within those categories. METR retains specified rights to comment on the access it received. Both parts matter. This isn't a claim that METR's work is compromised. It's a reason to ask how disputes over redactions get resolved when the company and evaluator disagree. Should an independent party settle those disputes? Editorial digest · November 15, 2026. Commentary based on the sources below; not a personal user account. Sources: https://www.anthropic.com/news/an-update-on-independent-evaluators