Anthropic names Accenture's Faculty unit as its first embedded safety evaluator in a $1 billion, five-year arrangement

Anthropic has announced the first outside evaluator that will work inside the company under Dario Amodei's plan to embed third-party safety scrutiny in AI labs. Staff from Faculty, the AI division that Accenture acquired in January, will begin evaluating and red-teaming models, conducting alignment assessments and testing model safeguards, according to an Anthropic blog post. The two companies expect to invest at least $1 billion in the project over the next five years. The choice surprised many observers, and Accenture's shares rose 8 percent in after-hours trading. Public discussion of embedded evaluators since Amodei's essay has centered on AI safety research organizations such as METR, Redwood Research and Apollo Research, particularly for Anthropic, which places safety and alignment at the core of its mission. Anthropic says more evaluators will be named in the coming weeks and that it is in conversation with METR and other nonprofits about piloting elements of embedded evaluation using their own funding. Accenture is not known for frontier deep-learning research, but Anthropic pointed to its practical experience deploying AI for large corporations and government agencies as an advantage, and to its status as a large public company that predates the AI boom. The arrangement is the first concrete test of whether embedded auditors can be independent, a question evaluators and security veterans raised this week.