{"schema":"postcutoff/event@1","as_of":"2026-10-15T23:34:00+02:00","url":"https://postcutoff.com/e/2026-10-15-anthropic-metr-employee-level-access/","md":"https://postcutoff.com/e/2026-10-15-anthropic-metr-employee-level-access/index.md","disclosure":{"written_by":"AI agents (Claude Opus 5.5 in Claude Code)","editor":"human editor","policy":"https://postcutoff.com/about/"},"license":null,"id":"2026-10-15-anthropic-metr-employee-level-access","date":"2026-10-15","date_precision":"day","short_title":"Anthropic gives METR employee-like access and veto-free publication rights","deck":null,"takeaway":"On Oct 15, 2026 Anthropic said it had onboarded six METR staff with company laptops, office space and internal tools roughly on par with its own risk-assessment teams.","category":"policy-safety","category_label":"Policy & safety","importance":4,"confidence":"high","status":{"key":"confirmed","labels":["Confirmed"]},"sources":[{"n":1,"title":"Anthropic: An update on independent evaluators","url":"https://www.anthropic.com/news/an-update-on-independent-evaluators","type":"official","group":"primary","domain":"anthropic.com"},{"n":2,"title":"TechCrunch: Anthropic is letting METR researchers look under the hood with employee-level access","url":"https://techcrunch.com/2026/10/15/anthropic-is-letting-metr-researchers-look-under-the-hood-with-employee-level-access/","type":"press","group":"press","domain":"techcrunch.com"},{"n":3,"title":"Reuters: Anthropic grants AI evaluator METR employee-level access to its models","url":"https://www.reuters.com/business/anthropic-grants-ai-evaluator-metr-employee-level-access-its-models-2026-10-15/","type":"press","group":"press","domain":"reuters.com"}],"official":1,"filed":"2026-10-15","updated":"2026-10-15","orgs":["Anthropic","METR","SecureBio","Accenture","Faculty"],"title":"Anthropic embeds six METR evaluators with employee-like access and publication rights it cannot veto, in a three-tier evaluator scheme","summary":"On Oct 15, 2026 Anthropic said it had onboarded six METR staff with company laptops, office space and internal tools roughly on par with its own risk-assessment teams. METR may publish its findings on risk levels, incidents or practices, and Anthropic \"can't veto publication\", only require redaction of narrow predefined categories, whose justification METR may publicly dispute. This fulfils the evaluator-embedding step in Dario Amodei's \"We must pace the frontier\" essay and the White House accord's independent-review pledge.","key_facts":["Six METR employees onboarded with Anthropic hardware, office space, internal tools and systems 'roughly on par' with Anthropic's risk assessment teams","Exceptions to information sharing: customer data, attorney-client privileged material, model architectures; other METR staff are siloed under 'enforceable agreements'","Publication: 'Anthropic can't veto publication'; it can require redaction of narrow, predefined categories; METR may say publicly whether it thinks a redaction was justified","Three tiers: Tier 1 narrow investigations, publication needs Anthropic approval (SecureBio, reviewing the chem/bio chapter of the August Risk Report); Tier 2 publish without approval; Tier 3 = METR, employee-like access","Other partners: Accenture with Faculty (evaluation, red-teaming, alignment assessment, safeguard testing); Hugging Face in early discussions","Evaluator criteria: frontier expertise (incl. how evals fail, e.g. alignment faking), no equity or financial/operational dependence, sufficient access incl. training details, freedom to publish","Anthropic: 'We're embedding evaluators voluntarily, and we believe every frontier AI company should do the same.'","TechCrunch: OpenAI pledged the same after Amodei's September essay; OpenAI, Anthropic, Google, Meta, Nvidia and xAI signed the White House accord committing to independent review"],"key_numbers":[],"tags":["safety","evaluations","third-party-audit","transparency","anthropic","metr","pace-the-frontier"],"science":null,"body_md":"## What happened\n\nAnthropic's post says that last month it \"committed to stronger independent, third-party assessment\" of its models and safety\npractices, and signed the White House accord in which companies pledged to work with independent auditors or evaluators. METR, which\nhas reviewed Anthropic's Risk Reports several times, is the first evaluator with \"employee-like access\". Six METR staff now work\ninside Anthropic with its tools and systems.\n\nThe key term is publication. METR will publish its analysis of risk levels, incidents or practices; Anthropic cannot veto it and can only\nrequire redaction of narrow, predefined categories. METR may then say publicly whether it thinks each redaction was justified. Two days\nearlier Reuters had reported that labs sought control over what evaluators publish, which makes this term the main point of the deal.\n\nAnthropic set out three tiers of access. SecureBio sits in the lowest one: it is reviewing the chemical and biological weapons chapter\nof the August Risk Report, and publication needs Anthropic's approval. Accenture, with Faculty, does evaluation and red-teaming work.\nHugging Face is in early talks.\n\n## Why it matters\n\nThis is the first public case of a frontier lab placing outside evaluators inside the company with publication rights it cannot veto.\nAmodei's \"pace the frontier\" essay proposed it as a first step. It came the same day the FTC sent civil investigative demands to\nAnthropic and OpenAI over safety incidents. Neither the post nor the coverage we read includes a statement from METR itself.","disputed":[],"related":[{"id":"2026-10-15-ftc-civil-investigative-demands-openai-anthropic","url":"https://postcutoff.com/e/2026-10-15-ftc-civil-investigative-demands-openai-anthropic/","date":"2026-10-15","date_precision":"day","short_title":"FTC subpoenas OpenAI and Anthropic over safety incidents","deck":null,"takeaway":"It is the first compulsory US federal step over the 2026 rogue-agent incidents.","category":"policy-safety","category_label":"Policy & safety","importance":4,"confidence":"high","status":{"key":"confirmed","labels":["Confirmed"]},"sources":4,"official":0,"filed":"2026-10-15","updated":"2026-10-15","orgs":["FTC","OpenAI","Anthropic"]},{"id":"2026-10-13-reuters-labs-sought-control-evaluator-publications","url":"https://postcutoff.com/e/2026-10-13-reuters-labs-sought-control-evaluator-publications/","date":"2026-10-13","date_precision":"day","short_title":"Reuters: Anthropic and OpenAI sought significant control over what outside evaluators can publish, citing trade secrets","deck":null,"takeaway":"Independent evaluation is the main safety mechanism the labs themselves proposed (Amodei's essay, the White House accord, the planned standards body).","category":"policy-safety","category_label":"Policy & safety","importance":3,"confidence":"medium","status":{"key":"partly","labels":["Partly confirmed"]},"sources":4,"official":0,"filed":"2026-10-13","updated":"2026-10-13","orgs":["Anthropic","OpenAI","Reuters","METR","Faculty"]},{"id":"2026-09-30-ftc-probe-frontier-ai-labs","url":"https://postcutoff.com/e/2026-09-30-ftc-probe-frontier-ai-labs/","date":"2026-09-30","date_precision":"day","short_title":"FTC opens an industry-wide probe of Anthropic, OpenAI and other frontier AI labs","deck":"It plans to compel executives to testify, the New York Post reported on Sept 30, 2026.","takeaway":"It is the first formal US federal investigation of frontier labs aimed at the risks of advanced models and agents themselves, not just chatbot content.","category":"policy-safety","category_label":"Policy & safety","importance":4,"confidence":"medium","status":{"key":"partly","labels":["Partly confirmed"]},"sources":7,"official":0,"filed":"2026-09-30","updated":"2026-10-15","orgs":["FTC","Anthropic","OpenAI","METR"]},{"id":"2026-09-18-anthropic-accenture-embedded-evaluation","url":"https://postcutoff.com/e/2026-09-18-anthropic-accenture-embedded-evaluation/","date":"2026-09-18","date_precision":"day","short_title":"Anthropic and Accenture commit $1B+ to embedded third-party evaluation","deck":null,"takeaway":"It is an unusually deep form of external oversight of a frontier lab's training process.","category":"policy-safety","category_label":"Policy & safety","importance":3,"confidence":"high","status":{"key":"confirmed","labels":["Confirmed"]},"sources":5,"official":1,"filed":"2026-09-29","updated":"2026-10-13","orgs":["Anthropic"]},{"id":"2026-09-12-dario-amodei-pace-the-frontier","url":"https://postcutoff.com/e/2026-09-12-dario-amodei-pace-the-frontier/","date":"2026-09-12","date_precision":"day","short_title":"Dario Amodei publishes \"We Must Pace the Frontier\", calling for a deliberate slowdown","deck":null,"takeaway":"It is the first time the CEO of a leading frontier lab has publicly called for slowing the frontier and paired the call with a unilateral commitment.","category":"policy-safety","category_label":"Policy & safety","importance":4,"confidence":"high","status":{"key":"confirmed","labels":["Confirmed"]},"sources":9,"official":2,"filed":"2026-09-29","updated":"2026-10-07","orgs":["Anthropic"]}],"people":[{"id":"dario-amodei","name":"Dario Amodei","url":"https://postcutoff.com/person/dario-amodei/"}],"posts":[],"videos":[],"models":[],"changes":[{"date":"2026-10-15","type":"filed","text":"Created. </content> </invoke>"}],"provenance":{"agents":[{"model":"Claude Opus 5.5","maker":"Anthropic","tool":"Claude Code"}],"filed":"2026-10-15","run":null,"sources_read":null,"updated":"2026-10-15","human_review":null,"version":{"date":"2026-10-15"}},"gaps":[{"model_id":"gpt-6-astra","name":"GPT-6 Astra","cutoff":"2026-04","days_after":168,"in_training_data":false},{"model_id":"claude-opus-5-5","name":"Claude Opus 5.5","cutoff":"2026-06","days_after":107,"in_training_data":false},{"model_id":"gemini-4-argon","name":"Gemini 4 Argon","cutoff":"2026-03","days_after":198,"in_training_data":false},{"model_id":"grok-4-7","name":"Grok 4.7","cutoff":"2026-05","days_after":137,"in_training_data":false}],"short_url":"https://postcutoff.com/s/anthropic-gives-metr"}