Multi-year deal
Magic Circle law firm Freshfields has teamed up with AI giant Anthropic to develop specialist legal tools.
Anthropic, the company behind the Claude family of AI models, will work with Freshfields to build tools to support the firm’s lawyers, with key areas of focus including legal and market research, contract review, document drafting and due diligence.
Under the multi-year deal, Freshfields will receive early access to new Anthropic models and tools and will roll out Claude across all 33 of its global offices. The agreement extends an existing partnership — Freshfields was already a client of Anthropic — and within the first six weeks of deploying Claude to its 5,700 employees, usage had already increased by around 500%, according to the firm.
The deal will also involve the development of new AI workflows to deliver legal services, as well as a focus on legal-oriented agentic tools, which will include Cowork, Anthropic’s agentic AI platform.
Gil Perez, chief innovation officer at Freshfields said:
“For us, innovation is about practical impact — helping our teams deliver the very best outcomes for clients, while holding ourselves to the highest standards of responsibility and governance. Partnering with Anthropic strengthens our ability to co-innovate at pace and to bring new capabilities into our work in a way that is secure, compliant and focused on client needs.”
Anthropic, currently valued at £280 billion following a $22 billion funding round in February 2026, has begun preparations for a potential IPO. Notably, the company’s reach into the legal sector extends beyond this deal. Several major legal AI platforms, including Harvey and Legora, already use Claude as the large language model underpinning their own systems.
The partnership comes after Freshfields announced the axing of a number of paralegal roles at its Manchester support hub. At the time, Freshfields cited the firm’s investment in technology and meeting client needs as the reasons for the cuts.
Having previously worked at Freshfields as a Principal Developer, this development is fascinating to watch. It reflects a broader industry transition from isolated AI experimentation toward fully integrated legal workflow infrastructure. The real opportunity is not just in deploying large language models, but in combining institutional legal knowledge, governance, retrieval systems, verification layers, workflow orchestration, and controlling operational intelligence into scalable operating models for legal services. Partnerships like this show that legal engineering is rapidly becoming a core strategic capability for global firms, particularly around areas such as programmable policy, compliance automation, and AI-assisted legal reasoning.