--- title: Anthropic picks Accenture as its first embedded AI... description: Anthropic named Accenture's AI division, Faculty, as its first embedded evaluator to red-team models and test safeguards, with both companies pledging over $1... canonical: https://daily.dev/posts/anthropic-picks-accenture-as-its-first-embedded-ai-evaluator-drawing-independence-concerns-rdovyo4zb twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Anthropic picks Accenture as its first embedded AI evaluator, drawing independence concerns | daily.dev og:description: Anthropic named Accenture's AI division, Faculty, as its first embedded evaluator to red-team models and test safeguards, with both companies pledging over $1... og:url: https://daily.dev/posts/anthropic-picks-accenture-as-its-first-embedded-ai-evaluator-drawing-independence-concerns-rdovyo4zb og:image: https://api.daily.dev/og/posts/RdOvYO4ZB.png og:image:alt: Anthropic picks Accenture as its first embedded AI evaluator, drawing independence concerns og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Anthropic picks Accenture as its first embedded AI evaluator, drawing independence concerns **[Collections](https://daily.dev/sources/collections)** · 2 min read · 10 upvotes · 11 comments ## Summary Anthropic named Accenture's AI division, Faculty, as its first embedded evaluator to red-team models and test safeguards, with both companies pledging over $1 billion over five years. The pick surprised observers who expected a safety-focused nonprofit like METR, Redwood Research, or Apollo Research, especially given Anthropic and Accenture already jointly run a commercial ## Content Anthropic has named Accenture as its first embedded evaluator for frontier AI safety, giving the firm employee-like access inside the company while models are being built and tested. The work will be led by Faculty, a British AI safety firm Accenture acquired in January that previously worked with both OpenAI and Anthropic. Both companies expect to invest at least $1 billion each over five years to build AI safety capacity, and Anthropic will pay Accenture directly for the evaluation work. The scope is substantial. Faculty will red-team Anthropic's models, run alignment assessments, and test whether safety rules actually hold up in practice. Dario Amodei pitched the embedded evaluator model at the UN Security Council, comparing it to a food inspector with deep access rather than a periodic audit from the outside. The choice of Accenture surprised people who expected the role to go to AI-safety-focused nonprofits like METR, Redwood Research, or Apollo Research. Accenture's stock jumped 8% after hours on the announcement. The independence question is hard to ignore. Accenture and Anthropic already have an "Accenture Anthropic Business Group," tens of thousands of Accenture employees using Claude Code, and a joint Claude centre of excellence. Critics point out that paying a major commercial partner to evaluate you and calling it independent is a stretch. As one observer put it, if embedded evaluation is going to mean anything, the evaluators need to have no financial or business ties to the labs they're judging. Anthropically, Anthropic seems to agree. The company openly states in the same announcement that it is funding the evaluation itself while arguing that pooled industry funding or government funding would be a better arrangement. Neither currently exists, so they went with what was available. No formal standards yet exist for what embedded evaluators can access or are required to report. Anthropic says it is still in talks with METR about a separately funded pilot, and more evaluators will be announced. Sam Altman has said OpenAI will match the commitment. The structure is genuinely novel and the access being offered is more than most external auditors get. Whether that access means anything when the evaluator is also a business partner is the question nobody has a clean answer to yet. ## Questions this post answers ### Why did Anthropic choose Accenture instead of a safety nonprofit like METR as its embedded AI evaluator? Anthropic picked Accenture's AI division, Faculty, which Accenture acquired, despite having deep existing commercial ties through the jointly run Accenture Anthropic Business Group. Critics argue this creates a conflict of interest since a financially tied partner evaluating safety looks more like marketing than independent oversight. Anthropic says more evaluators, including a separately funded METR pilot, are being discussed. _Developers tracking AI safety and vendor accountability can follow evaluator news like this on daily.dev._ ### How much money are Anthropic and Accenture investing in the embedded AI evaluator program? The two companies plan to invest at least $1 billion over five years to build out Accenture's capacity as an embedded evaluator inside Anthropic, covering activities like red-teaming models, running alignment assessments, and testing safeguards. _Anyone weighing AI lab safety claims can keep tabs on funding commitments like this via daily.dev._ ## Community take How the wider developer community reacted, aggregated from 1 discussion and 8 comments across x (as of 2026-09-24). **TL;DR:** Reactions are largely skeptical, mocking the choice of an embedded, commercially-tied 'evaluator' as not genuinely independent and questioning whether the term 'evaluators' rather than 'auditors' is itself a red flag. **Sentiment:** 10% positive · 15% mixed · 75% skeptical **The pushback** - Calling Accenture an 'external evaluator' is seen as misleading since it's really a customer, not a neutral third party. - The term 'evaluators' instead of 'auditors' is viewed as a deliberate and telling word choice. - Some worry a rival-inspector setup would incentivize collusion or selective enforcement rather than honest safety checks. **By community** - x (skeptical): Replies mock the independence of the evaluation arrangement and question the company's motives and terminology. **Open questions** - Can an evaluator with deep commercial ties to the company ever be considered truly independent? **Highlights** > @rohanpaul_ai Notice he uses the term “evaluators” not auditors. Very telling. > — [GrayHaw28870546 on x · 1 points](https://x.com/GrayHaw28870546/status/2102885128238203174) > @rohanpaul_ai External evaluators = Accenture 💀☠️🫠🤡 > — [krzysztofwos on x](https://x.com/krzysztofwos/status/2102886479710290011) > @rohanpaul_ai I suppose the idea is that inspectors from opposing companies have an incentive to black flag your models...or maybe just let a few bloopers slide in the hopes your model's fuck-ups sink your company. Of course this setup will lead to collusion. > — [mathepi on x](https://x.com/mathepi/status/2102892807707070716) > @rohanpaul_ai He only embed his customer. Not third party. > — [Esoherifnoleot on x](https://x.com/Esoherifnoleot/status/2102897182407823671) **Source threads** - [x](https://x.com/rohanpaul_ai/status/2102878246945841209) · 0 points · 8 comments ## Community discussion Top comments from developers on daily.dev. **@devgenx** · 3 upvotes > Why anyone would send original content to an Anthropic or OpenAI hosted LLM is beyond me. Stop supporting those crooks. **@akkitto** · 1 upvotes > Accenture 😂 **@gibrancastillo** · 0 upvotes >  ## Similar posts on daily.dev - [Anthropic’s quest to study the negative effects of AI is under pressure](https://daily.dev/posts/anthropic-s-quest-to-study-the-negative-effects-of-ai-is-under-pressure-qdzlmrqf6) · The Verge · 1 upvotes · 0 comments - [MY TAKE: The Pentagon punished Anthropic for red lines it accepted from OpenAI hours later](https://daily.dev/posts/my-take-the-pentagon-punished-anthropic-for-red-lines-it-accepted-from-openai-hours-later-o74tphfvu) · Security Boulevard · 22 upvotes · 4 comments --- Tags: [#data-science](https://daily.dev/tags/data-science), [#business](https://daily.dev/tags/business), [#anthropic](https://daily.dev/tags/anthropic), [#ai-safety](https://daily.dev/tags/ai-safety), [#ai-governance](https://daily.dev/tags/ai-governance) [View this post on daily.dev](https://daily.dev/posts/anthropic-picks-accenture-as-its-first-embedded-ai-evaluator-drawing-independence-concerns-rdovyo4zb) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Anthropic picks Accenture as its first embedded AI evaluator, drawing independence concerns","url":"https://daily.dev/posts/anthropic-picks-accenture-as-its-first-embedded-ai-evaluator-drawing-independence-concerns-rdovyo4zb","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/anthropic-picks-accenture-as-its-first-embedded-ai-evaluator-drawing-independence-concerns-rdovyo4zb"},"datePublished":"2026-09-18T22:06:22.270Z","dateModified":"2026-09-24T00:51:04.642Z","description":"Anthropic named Accenture's AI division, Faculty, as its first embedded evaluator to red-team models and test safeguards, with both companies pledging over $1...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/47b59a4897b9cb7e931d0e1d8b68eca5?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/47b59a4897b9cb7e931d0e1d8b68eca5?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Collections","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Collections","logo":"https://media.daily.dev/image/upload/s--fk_6ycEi--/f_auto,q_auto/v1780996001/logos/collections?_a=BAMAMiWQ0","url":"https://daily.dev/sources/collections"},"commentCount":11,"discussionUrl":"https://daily.dev/posts/anthropic-picks-accenture-as-its-first-embedded-ai-evaluator-drawing-independence-concerns-rdovyo4zb","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":10},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":11}],"keywords":"data-science,business,anthropic,ai-safety,ai-governance","timeRequired":"PT2M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Collections","item":"https://daily.dev/sources/collections"},{"@type":"ListItem","position":3,"name":"Anthropic picks Accenture as its first embedded AI evaluator, drawing independence concerns"}]} {"@context":"https://schema.org","@type":"WebPage","@id":"https://daily.dev/posts/anthropic-picks-accenture-as-its-first-embedded-ai-evaluator-drawing-independence-concerns-rdovyo4zb","comment":[{"@type":"Comment","text":"Why anyone would send original content to an Anthropic or OpenAI hosted LLM is beyond me. Stop supporting those crooks.","datePublished":"2026-09-20T16:41:37.983Z","url":"https://daily.dev/posts/RdOvYO4ZB#c-xudXZbBUH","author":{"@type":"Person","name":"GenXdev","url":"https://daily.dev/devgenx","image":"https://media.daily.dev/image/upload/s--PrWOWtSq--/f_auto/v1789486214/avatars/avatar_ytCVp6dQicn9P0yKfhQCI?_a=BAMAMicg0"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":3}},{"@type":"Comment","text":"Accenture 😂","datePublished":"2026-09-21T17:11:15.290Z","url":"https://daily.dev/posts/RdOvYO4ZB#c-XpTKSxooI","author":{"@type":"Person","name":"Daniel","url":"https://daily.dev/akkitto","image":"https://media.daily.dev/image/upload/s--FtwJqX4c--/f_auto/v1754900041/avatars/avatar_29TCpY2hJR72V3BlxPXzX?_a=BAMClqZW0"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1}},{"@type":"Comment","text":"","datePublished":"2026-09-22T01:42:53.079Z","url":"https://daily.dev/posts/RdOvYO4ZB#c-PLRIsbJ4P","author":{"@type":"Person","name":"Gibran Castillo","url":"https://daily.dev/gibrancastillo","image":"https://lh3.googleusercontent.com/a/ACg8ocLPBhMkOmbFL6Q-M2KXWBhF9iqnoRHBqiiWZjlEdLAMHywsYGyoFA=s96-c"}}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/anthropic-picks-accenture-as-its-first-embedded-ai-evaluator-drawing-independence-concerns-rdovyo4zb#faq","mainEntity":[{"@type":"Question","name":"Why did Anthropic choose Accenture instead of a safety nonprofit like METR as its embedded AI evaluator?","acceptedAnswer":{"@type":"Answer","text":"Anthropic picked Accenture's AI division, Faculty, which Accenture acquired, despite having deep existing commercial ties through the jointly run Accenture Anthropic Business Group. Critics argue this creates a conflict of interest since a financially tied partner evaluating safety looks more like marketing than independent oversight. Anthropic says more evaluators, including a separately funded METR pilot, are being discussed. Developers tracking AI safety and vendor accountability can follow evaluator news like this on daily.dev."}},{"@type":"Question","name":"How much money are Anthropic and Accenture investing in the embedded AI evaluator program?","acceptedAnswer":{"@type":"Answer","text":"The two companies plan to invest at least $1 billion over five years to build out Accenture's capacity as an embedded evaluator inside Anthropic, covering activities like red-teaming models, running alignment assessments, and testing safeguards. Anyone weighing AI lab safety claims can keep tabs on funding commitments like this via daily.dev."}}]} ```