Anthropic embeds Accenture evaluators, $1B each
Anthropic will give Accenture's AI unit Faculty employee-level access to evaluate and red-team its frontier models, the first concrete step in a commitment to embed outside evaluators inside the company.
Anthropic said it is partnering with Accenture on independent evaluation of its frontier models, delivering on a pledge made in CEO Dario Amodei's essay "We Must Pace the Frontier" to embed evaluators within the company.
The work will be led by Faculty, Accenture's specialist AI business, and will cover model evaluation and red-teaming, alignment assessments, and testing of model safeguards. Anthropic and Accenture each expect to invest at least $1 billion in building capacity over the next five years.
The distinguishing feature is access. Unlike external evaluators, embedded evaluators will work inside the company with access "comparable to an employee's" — observing models during training, following deployment decisions, and speaking directly to staff. Anthropic says they will be able to verify safety commitments, identify blind spots, and report incidents publicly, while stressing that "the safety of our models remains our responsibility."
Anthropic is candid that the arrangement is unsettled. There are no standards for what information embedded evaluators should access or how they should report findings, and no established funding system. Anthropic says funding should eventually come from pooled or government sources, as proposed in its Advanced AI Framework in June; absent that, it will fund Accenture's work directly. It is separately in talks with METR and other nonprofit evaluators to pilot embedded evaluation using their own funding.
The partnership is non-exclusive. Anthropic says additional evaluators will be announced in the coming weeks. No evaluation results have been published.