Accenture and Anthropic partner to build embedded AI safety evaluator team

NEW YORK: Accenture and Anthropic have partnered to establish a team of embedded evaluators that will work alongside Anthropic’s internal teams and safety partners to test and assess its AI models.

The companies said the team will evaluate and red-team models, conduct alignment assessments and test safeguards as AI systems continue to become more capable and widely deployed.

As part of the partnership, both Accenture and Anthropic expect to invest at least $1 billion each in AI safety over the next five years.

Accenture will draw on its investment in responsible AI and technical safety expertise through Faculty, the applied AI company it acquired.

Faculty has experience testing and evaluating models for leading AI labs and developing AI systems designed with safety and ethics in mind. Its work spans areas including government, defence, healthcare and infrastructure.

The company was also involved in developing the UK National Health Service’s Early Warning System during the Covid-19 pandemic.

“Accenture is bringing together a dedicated team with deep AI, security and industry expertise to work alongside Anthropic,” Accenture chair and CEO Julie Sweet 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,” she added.

Accenture chief technology officer and Faculty CEO Dr Marc Warner said the partnership builds on Faculty’s focus on developing AI systems with safety considerations built in from the start.

“Faculty, which is now part of Accenture, was founded on the belief that AI should be safe by design, not safe by accident,” Warner said.

He added that Faculty’s experience working with governments, defence and healthcare systems would support its role as an embedded evaluator alongside Anthropic.

The partnership comes as AI developers increasingly focus on model evaluation, red-teaming and safeguards as part of the development and deployment process. By placing external evaluators alongside Anthropic’s teams, the companies aim to bring additional technical and real-world expertise into the safety process.

Both companies’ planned investments of at least $1 billion over five years also underline the scale of their focus on AI safety as the technology moves into increasingly complex applications.