The team of individuals with AI, security and industry expertise will work with Anthropic’s internal teams and safety partners.

Anthropic and Accenture each expect to invest ‘at least $1bn’ over five years in AI safety, according to press releases issued by the companies.

Faculty, Accenture’s specialist AI business, will lead the partnership, which will include evaluating models, conducting alignment assessments, and testing model safeguards. The partnership will also involve ‘red-teaming', which is a process of trying to manipulate AI models before they are released.

Accenture acquired Faculty in early 2026. Its work includes government, defence and healthcare, including development of the UK National Health Service’s early warning system during the pandemic.

The partnership will consist of ‘embedded evaluation', as opposed to external valuation, meaning that embedded evaluators work inside AI companies, with access that allows them to see models take shape in training, track how decisions are made and speak to employees. 

Anthropic said independent embedded evaluators do not ‘reduce’ its accountability, but ‘make it more verifiable'.

Julie Sweet, chair and chief executive, Accenture, said. “Safety requires both deep technical expertise and a clear understanding of how AI is used in the real world. 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.”

Several AI industry insiders have shared warnings over AI’s risks in recent weeks. Anthropic safety researcher Evan Hubinger said that there is a 10% chance that AI could ‘kill all humans’ within the next decade, and OpenAI’s chief scientist, Jakub Pachocki, called for ‘extreme caution’ over the speed of AI development.