Anthropic named Accenture as its first embedded evaluator, tasked with observing model development from inside the company as part of a push to make frontier AI safety more independently verifiable. Anthropic and Accenture each plan to invest at least $1 billion over five years building this evaluation capacity, with Accenture’s Faculty unit conducting red-teaming and alignment assessments. Anthropic said it remains ultimately responsible for its models’ safety and is separately in discussions with the nonprofit METR and other potential evaluators.
