Anthropic said it is partnering with Accenture on independent evaluation of frontier AI, the first step in the embedded evaluator commitment its CEO described in 'We Must Pace the Frontier'; the work will be led by Accenture's AI unit, Faculty. Each side expects to invest $1 billion over five years to build capability in this area. The non-exclusive partnership covers evaluating and red-teaming models, conducting alignment assessments, and testing model safeguards. Anthropic said embedded evaluators will have employee-equivalent access to observe models during training, track build and deployment decisions, and speak directly with staff, while the arrangement does not reduce Anthropic's own responsibility for model safety.

Editor's note

This summary is based on the primary source. Open the original for full context, data, and qualifying details.