Anthropic and Accenture to commit at least $2 billion to AI safety
Anthropic and Accenture launched a partnership to create a team of embedded evaluators that will independently assess frontier AI models.
Each company expects to spend at least $1 billion over five years. Faculty, Accenture’s specialist AI business, will lead testing and assessment of Anthropic’s model safeguards.
Partnership model
On September 18, 2026, Anthropic and Accenture announced a pilot of embedded evaluation, in which independent evaluators work inside an AI laboratory. They are expected to receive access comparable to an employee’s into how models are trained and deployed.
Faculty, Accenture’s specialist AI business, will red-team models, conduct alignment assessments and test safeguards. The partnership is non-exclusive, so both parties may work with other evaluators and developers.
Funding and the independence question
Anthropic and Accenture each expect to spend at least $1 billion over the next five years. The combined commitment is at least $2 billion; it is not a one-time funding round into Anthropic. The announcements did not provide an annual spending schedule or headcount.
Anthropic acknowledges that common standards have not yet been established for evaluator access or publication of findings. Initially, Anthropic will fund Accenture’s work directly, raising questions about practical independence.
Sources and image credits
Sources
News date: September 18, 2026. Anthropic: https://www.anthropic.com/news/accenture-embedded-evaluation
Accenture: https://newsroom.accenture.com/news/2026/accenture-and-anthropic-partner-to-build-team-of-embedded-evaluators-at-anthropic
Reuters: https://www.reuters.com/business/anthropic-accenture-invest-2-billion-ai-model-evaluation-safety-concerns-rise-2026-09-18/
This article was prepared by the Founders Cafe editorial team.
