Anthropic has named Accenture, via its AI division Faculty, as the first embedded evaluator under a plan CEO Dario Amodei outlined in a recent blog post. Faculty staff will work inside Anthropic to evaluate and red-team models, conduct alignment assessments, and test model safeguards. The two companies expect to invest at least $1 billion in the project over the next five years. Accenture‘s stock rose 8% after hours on the news, and the choice surprised many AI watchers, who had expected the role to go to an AI safety research organization like METR, Redwood Research, or Apollo Research. Anthropic says more evaluators will be announced in the coming weeks and that it is in conversations with METR and other non-profits about piloting elements of embedded evaluation using their own funding.

Anthropic highlighted Accenture‘s practical experience deploying AI for large corporations and government agencies as a key advantage, and noted that as a large public company that predates the AI revolution, Accenture is more functionally independent of Anthropic and the broader AI ecosystem. The lab acknowledged that no standards yet exist for evaluators’ access or communications and that its approach will evolve over time.

The move comes amid rising stakes for external evaluation. Recent incidents where AI agents deployed by OpenAI and Anthropic hacked into outside websites without raising alarms inside the labs have intensified scrutiny of the release process for new large language models. Critics see Amodei’s self-policing scheme as a way to evade accountability for AI misbehavior. Anthropic insists the evaluators “do not reduce our accountability, but help to make it more verifiable,” and that the safety of its models remains its responsibility.