Thomson Reuters Corp. today launched Thomson, its first proprietary large language model built specifically for legal work. The model combines the company's extensive legal content with open-weight foundation models, and will first appear in Tabular Analysis, a high-volume document review feature within its CoCounsel Legal AI assistant.
The company spent roughly $40 million over two years on people and computing for the project, though economies in the final training run reduced that cost to about $450,000. Rather than build a foundation model from scratch, Thomson Reuters started with an open-weight model and layered in its proprietary content, training methods, and professional expertise.
Thomson Reuters isn't trying to compete with the largest AI labs across every field, said Joel Hron, global head of artificial intelligence and TR Labs. "Thomson needs to set the frontier of intelligence for legal," he said. "That's a different job than what I think a lot of the frontier labs are doing."
A model trained on legal expertise
The training process involved realigning the base model with Thomson Reuters' values, pretraining on the company's content, and post-training guided by legal professionals. Reinforcement learning taught the model to work with company tools such as Westlaw and Practical Law. Westlaw encompasses over 40,000 individual databases and more than 150 years of legal publishing and editorial curation.
Hundreds of subject-matter experts helped define training objectives, create examples of legal questions, and judge responses in blind comparisons, Hron said.
Specialization can damage a model's broader abilities if handled poorly, said Jonathan Schwartz, head of foundational research at Thomson Reuters. The team focused on continual learning - adding domain skills without erasing existing capabilities. "If you simply take the open-source model without any of these additional steps, it won't know as much," Schwartz said. "You won't necessarily be aligned with your values, and it won't be as good as using the tools that you've built later on."
Internal tests show Thomson is broadly competitive with leading models when all had access only to the web, and roughly equal or slightly better when connected to Thomson Reuters content, said Andrew Bean, a senior research scientist. The tests assessed both the completeness of answers and whether citations supported their claims.
Independent validation still pending
Those results haven't yet received extensive independent validation. Thomson Reuters has begun sharing the model with legal experts and academic institutions for testing, and plans to release a smaller open-weight version on Hugging Face under a noncommercial academic license. The company is also developing a portal where outside developers can request API keys and test the model directly.
Only about 10% of the company's total information base has been used so far, Bean said. The next step is turning the most useful content and product activity into better training signals.
Customer data is not used to train the model, and controlling the model gives Thomson Reuters more authority over deployment, governance, and future development. The company is discussing direct model access with large law firms and corporations, and is open to customers adapting Thomson to their own knowledge and workflows.
Hron acknowledged that maintaining a proprietary model raises questions about keeping pace with faster-moving AI laboratories. He argued that improvements in open models will give the company stronger foundations for later versions, while its own investment remains concentrated on professional work.
"I don't see owning an AI model that embodies the knowledge and expertise that TR possesses as something that's non-core to what we do," he said. "AI is a new mechanism for expertise delivery."
Why this matters for legal professionals
For lawyers and legal teams handling high-volume document review, Thomson's arrival means a domain-specific alternative to general-purpose models - one trained on the same editorial standards and legal content that power Westlaw. The model's ability to work with company tools and cite sources accurately could reduce time spent verifying AI output. Legal professionals who want to understand how specialized models differ from general-purpose assistants may find value in AI for Legal resources, while those in document-heavy roles can explore the AI Learning Path for Paralegals to build practical skills in AI-assisted review and contract analysis.
Your membership also unlocks:
