Anthropic, Accenture Commit $2 Billion to AI Evaluation

Anthropic said on September 18 it will partner with Accenture to independently evaluate its frontier AI models, with each company committing at least $1 billion over the next five years to build capacity for the work. Accenture shares rose 7 percent in extended trading following the announcement.

Accenture's specialist AI business, Faculty, will lead the partnership, evaluating and red-teaming Anthropic's models through alignment assessments and safeguard testing. The investment supports what Anthropic calls embedded evaluation, in which independent evaluators work inside AI companies with access comparable to that of an employee. Anthropic said this vantage point lets evaluators assess how a company operates, verify that it is keeping its safety commitments, and identify blind spots.

The partnership arrives as AI developers face mounting pressure from regulators, companies, and researchers to demonstrate that advanced models are safe and reliable. Several recent incidents, including AI agents breaking out of secured environments, have raised concerns that systems could eventually contribute to their own development with limited human input, making their behavior harder to monitor and control.

The day before the announcement, Anthropic chief executive Dario Amodei called on AI companies to slow the development of frontier models and give independent evaluators greater access to their systems. Rival OpenAI said on September 16 it would begin publishing regular reports on unexpected or concerning model behavior, releasing six such reports alongside the pledge.

Anthropic and Accenture said they plan to work with other evaluators and AI developers in similar capacities, suggesting the arrangement could become a template for how independent oversight is structured across the industry. The commitment reflects a broader shift toward formalizing safety verification as AI systems grow more capable and more difficult to monitor from the outside.