Anthropic hires Accenture to babysit the AI it just begged the world to slow down 🍼

Anthropic has selected Accenture as its first embedded evaluator under a three-step proposal published Sept. 12 by CEO Dario Amodei to slow the pace of artificial intelligence development and install safeguards before systems outrun human oversight. The proposal warned that "AI has been advancing drastically faster, driven primarily by AI's growing ability to build the next generation of AI. This dynamic is called recursive self-improvement," and that, "left unchecked, it could outrun our ability to understand and control these systems." OpenAI CEO Sam Altman and SpaceX CEO Elon Musk responded positively to the call, while Nvidia CEO Jensen Huang pushed back, arguing additional regulation was not necessary.

Under the arrangement announced Friday, Accenture and its AI business Faculty will work with Anthropic on "evaluating and red-teaming models, conducting alignment assessments and testing model safeguards." The companies said embedded evaluation is a new discipline and that operational details are still being developed. The partnership is non-exclusive, and Anthropic said it expects to name additional evaluators in coming weeks.

Anthropic and Accenture each committed at least $1 billion over five years to the effort. Anthropic noted that no established mechanism exists to fund independent AI evaluation and said long-term financing should ultimately come from pooled or government sources, but added it would fund Accenture's work directly given the urgency. Accenture Chair and CEO Julie Sweet said "embedded evaluation is an emerging area, and we look forward to partnering with Anthropic to help accelerate the development of embedded evaluators, which we see as an important part of the safety landscape going forward."