The legal sector, once feared to be disrupted drastically by AI, is now embracing large language model (LLM) technology through strategic collaborations between law firms and AI developers. Freshfields and Anthropic have launched a co-innovation program to build and deploy agentic AI workflows that enhance legal service delivery across global offices.
- Freshfields and Anthropic co-develop AI workflows to enhance legal services
- Focus on building agentic AI to handle complex, multi-step legal tasks
- Thomson Reuters targets ‘fiduciary-grade AI’ for regulated legal and financial work
What happened
Freshfields, an international law firm, has partnered with Anthropic to integrate and co-develop large language model (LLM) based workflows tailored for legal practice. Their joint initiative focuses on creating AI-native legal services by embedding Anthropic’s Claude LLM technology into Freshfields’ proprietary AI platforms, currently serving thousands of users globally.
This collaboration is not just about tool adoption but involves co-innovation, where in-house legal and AI teams work closely to design agentic workflows capable of managing end-to-end, multi-step legal tasks. The initiative aims to provide lawyers with secure, efficient AI-powered processes that leverage institutional knowledge to support client work across all Freshfields offices.
Why it matters
The legal industry has long been cautious about AI replacing professional legal work due to the high stakes and complexity involved. By focusing on co-developed, domain-specific AI solutions, firms like Freshfields are addressing these concerns directly, enabling AI to complement rather than replace human expertise. This reduces risks related to errors and trust, which are critical in legal and regulatory environments.
Similarly, Thomson Reuters is advancing its legal AI offerings with ‘fiduciary-grade AI,’ specifically designed to meet the rigorous demands of law, tax, and audit tasks where errors have severe consequences. This approach seeks to create deterministic AI tools whose outputs legal professionals can reliably verify and stand behind, contrasting with general-purpose AI models prone to probabilistic errors.
What to watch next
Over the next year, the collaboration between Freshfields and Anthropic will be a key development to monitor as their co-developed workflows are iterated and scaled for broader legal use. The success of these agentic AI applications could set new standards for AI integration in complex professional services and encourage wider adoption by other law firms.
Additionally, how Thomson Reuters progresses with its fiduciary-grade AI and integrates its proprietary LLM within legal and financial services will be important for the market. These developments will help define the competitive landscape for AI-powered legal technology and shape expectations around performance, compliance, and professional reliability.