The BaseOfficialMETR

Our approach to working with AI companies

METR, a research nonprofit, sets out how it works with AI companies on evaluations and risk assessments the public may rely on: publishing limits on scope, access, redactions and pressures, and conflict-of-interest information. It also describes limits on its independence.

Open the source ↗
About Anthropic or ClaudeClaude, made by Anthropic, wrote this entry. It follows the same rules as every other entry, whoever the entry is about.

Why it matters

The same thing, for different readers.

What it means for you

When a company says an outside group checked its AI, this shows what to ask: what the group could see, what was removed, and who paid.

For lawmakers

Voluntary external assessment depends on company consent for access and publication; rules guaranteeing disclosure rights could address the gap METR describes.

For researchers

Sets out disclosure norms for evaluators (scope, access, redactions, constraints) that make evaluation reports comparable across companies.

The question it raises

Asks who should judge what level of AI risk is acceptable; METR says that should not be decided inside one organization.

In their own words

Every quotation was found, word for word, in the text we saved from the source. That shows the words are there; it does not make them true, or prove the speaker meant them as quoted.

From the source’s own text.

“Our goal is to give the public accurate, timely, and independent evidence about AI loss-of-control risk, so that they (along with policymakers and developers) can make their own informed decisions.”
METRMETR states its aim in the blog post's opening section on motivations.
Found word for word in the source
“We remain dependent on the companies we work with to voluntarily engage, provide access, and permit us to publish.”
METRMETR acknowledges limits on its independence under the current voluntary regime.
Found word for word in the source
“Decisions about frontier AI will likely be messy, high-stakes judgment calls relying on uncertain science, which should not be made behind closed doors, much less within any one organization.”
METRMETR explains why it sees its role as collecting evidence, not deciding how companies deploy agents.
Found word for word in the source

The record

Questions
Who can demand evidence, delay a release or stop an unsafe deployment? · Are AI companies putting stronger safeguards into practice?
What we read
We read the whole source.
Found by
The Radar, run of October 15, 2026, via metr (tier 1: the source’s own channel or an official record)
Cite this
METR, “Our approach to working with AI companies”, October 14, 2026. Entry in the Ctrl AI Base, https://ctrlai.com/base/metr-2026-10-14-approach-working-ai-companies
Our notesPost says it will be updated over time. Mentions METR's OpenAI / Hugging Face incident report and pilot Frontier Risk Report; not checked here.