Thomson Reuters Launches Its First Proprietary AI Model, 'Thomson,' Built on Decades of Legal and Financial Data for Just $40 Million
Summary
Thomson Reuters launches 'Thomson,' its first proprietary AI model trained on decades of Westlaw, Reuters, and financial data, costing just $40 million to develop while reportedly matching frontier model performance — now live in CoCounsel Legal.
Key Points
- Thomson Reuters launches 'Thomson,' its first proprietary large language model, built on an open-source foundation and trained on decades of proprietary content from Westlaw, Practical Law, Checkpoint, and Reuters for use in professional industries.
- The model costs $40 million to develop — a fraction of what frontier labs spend — giving Thomson Reuters full control over its AI, reduced dependence on external roadmaps, and lower operational costs, with early evaluations showing performance on par with leading frontier models.
- Thomson is now available in CoCounsel Legal and raises questions about Thomson Reuters' future content licensing strategy, given its previous deals with AI labs like Meta and Microsoft.