Collaboration Announcement with Accenture for Integrated Evaluation
Anthropic announced on September 18, 2026, a collaboration with Accenture to conduct an independent integrated evaluation of advanced artificial intelligence. This partnership is part of Anthropic's commitment to incorporating internal evaluators into the development of its AI models.
Scope of the Collaboration with Accenture
The agreement entails that independent evaluators will work within Anthropic with a level of access comparable to that of an employee. This will allow them to directly observe the development and deployment of the models, follow decisions regarding their construction, and engage with employees. The collaboration includes evaluating and conducting red-teaming of models, performing alignment analyses, and testing the safety mechanisms of the models. Faculty, Accenture's specialized AI division, will lead these activities.
Committed Investment
Anthropic and Accenture plan to invest a significant amount each over the next five years to develop capabilities in this area. Anthropic will directly fund Accenture's work, as there is currently no established system for financing independent evaluations.
Current Status and Future of Integrated Evaluation
Integrated evaluation is a new practice, and many details about its operation are still under development. There are no defined standards regarding the information that integrated evaluators should have access to or how they should report their findings. In the long term, Anthropic believes that funding should come from pooled or governmental sources, although for now, it is assuming direct funding.
Collaboration with Other Evaluators and Future Model
The collaboration with Accenture is non-exclusive. Anthropic will work with other evaluators who will be announced in the coming weeks. Additionally, it is in discussions with METR and other non-profit organizations to pilot elements of integrated evaluation with its own funds. Anthropic expresses that frontier artificial intelligence needs an ecosystem of evaluators operating with shared standards and that laboratories should work with multiple organizations simultaneously.
Responsibilities and Accountability
Anthropic maintains responsibility for the safety of its models. The presence of integrated independent evaluators does not diminish this responsibility; rather, it contributes to making it more verifiable.
Published: 22/09/2026. Content reviewed using experience, authority and trustworthiness criteria (E-E-A-T).
This article may have used artificial intelligence tools to support structure, editing, translation or review. Editorial responsibility and final review remain with Toni Berraquero. View AI policy
You can support the project or share this article in one click. At least this block does something useful.