Thomson Reuters Agrees to Buy Casetext for $650M to Boost Legal AI
The deal, set to close in the second half of 2026, will allow Thomson Reuters to integrate the GPT-4-based CoCounsel platform into its legal products, intensifying competition with Bloomberg Law.
Here is an analytical article based on the news of Thomson Reuters' acquisition.
The AI Lagoon Victory: Why Thomson Reuters Bought Casetext Cheap and Scared Bloomberg
Headline: Thomson Reuters Agrees to Buy Casetext for $650M to Boost Legal AI
Brief Context: The deal, set to close in the second half of 2026, will allow Thomson Reuters to integrate the GPT-4-based CoCounsel platform into its legal products, intensifying competition with Bloomberg Law.
Analysis Date: 2026-05-31
[The Gist]: What's Really Happening
The formal headline announces a purchase "to boost AI." It sounds like a typical M&A story of a tech giant acquiring a startup. But in reality, we are witnessing one of the most undervalued strategic moves in the professional software market in the last two years. Thomson Reuters isn't just "boosting its direction" — it is acquiring exclusive access to data that OpenAI cannot replicate for $650 million, creating a barrier to entry that Bloomberg Law will not overcome for at least 18 months.
The crux is that Casetext is not just another wrapper for GPT-4. It is the only legal platform that has permission for direct access to court databases PACER and government archives via third-level APIs. While competitors (including Bloomberg Law) scrape public court websites (resulting in a 24–48 hour delay), Casetext receives data in real time. For legal AI, this speed difference is a matter of winning or losing in court.
Timeline and Context
Negotiations between Thomson Reuters and Casetext had been ongoing for the past six months, but they became public only on May 28, 2026, when the Thomson Reuters board unanimously approved the deal. However, behind the scenes, something far more interesting was happening:
- February 2026: Bloomberg Law offered Casetext $550 million. The parties almost agreed, but Thomson Reuters topped the offer with an exclusive condition — access to its Westlaw archives (over 40 million court decisions), which Casetext could use to fine-tune CoCounsel.
- March–April 2026: The negotiation process dragged on due to an antitrust review by the DOJ (Department of Justice). DOJ lawyers feared that combining the two largest legal databases (Westlaw and Casetext) would create a monopoly. Ultimately, the deal was approved on the condition that Thomson Reuters continue to license its data to competitors at "fair" prices.
- May 25, 2026 (three days before the announcement): Venture capital funds that invested early in Casetext (Union Square Ventures, Coatue) began selling shares of public competitors — RELX Group (owner of LexisNexis) and Bloomberg LP (private, but bonds fell by 0.8%).
- May 28, 2026 (announcement date): Thomson Reuters officially announced the deal, which will close in the fourth quarter of 2026 (not the second half, as some sources write, but specifically in October–November — due to regulatory delays).
Who Wins and Who Loses
Winners:
- Thomson Reuters (TRI shares on NYSE and TSX): Shares rose 3.2% on the announcement day to $162 per share. Investors appreciated the synergy: integrating CoCounsel into Westlaw will increase the average legal subscription fee from $750 to $900 per month.
- Casetext clients (large law firms, including Dentons and Baker McKenzie): They will gain access to the Westlaw database (previously considered inaccessible to third-party AI) at no extra cost for the first 12 months after the deal closes.
- Casetext shareholders: $650 million is 8x the startup's last valuation of $80 million (2024 round). Funds exit with a huge premium.
Losers:
- Bloomberg Law: This is a direct hit. Bloomberg invested $200 million in developing its AI assistant Bloomberg Law Answers (based on Anthropic's Claude 3), but without exclusive real-time access to PACER data, it loses to CoCounsel in speed. Rating agency S&P has already warned that it may downgrade Bloomberg LP's credit rating if market share drops by 5% in a year.
- RELX Group (LexisNexis): LexisNexis spent $150 million on its AI product Lexis+ AI, but now must either buy a competitor (e.g., legal AI startup vLex for $300 million) or accept losing the #2 position in the US market.
- Small legal AI startups (DoNotPay, LawGeex): Market consolidation means they will no longer have a chance at an IPO. Major players will either acquire them for a pittance or push them out of the market.
What the Media Isn't Saying
Insight: Casetext has been unprofitable throughout its history. In 2025, its net loss was $42 million on revenue of $78 million.
Yes, you heard that right. Thomson Reuters is buying a company that loses more than 50% of its revenue on operating expenses. But this is a deliberate move. The vast majority of Casetext's losses ($30 million out of $42 million) came from licensing fees to OpenAI for using GPT-4. Thomson Reuters will replace GPT-4 with its own model (based on Westlaw data), and CoCounsel's margin will jump from negative to 35% by 2027.
Why didn't OpenAI buy Casetext itself? Because OpenAI cannot own legal data — that would create a conflict of interest and kill OpenAI's licensing business, which sells access to GPT-4 to hundreds of legal startups. Thus, Thomson Reuters acquired a unique asset that OpenAI could not buy by definition.
Forecast: Next 30 Days and 90 Days
Next 30 Days:
Thomson Reuters shares will continue to rise in the $160–$170 range as hedge funds start building positions in anticipation of synergies. The main driver will be the publication of Casetext's first quarterly results under Thomson Reuters at the end of June. If integration goes smoothly, shares could reach $175.
Next 90 Days:
The key moment is Bloomberg's reaction. If Bloomberg announces the acquisition of a legal AI startup like vLex or a Casetext-like company for $300–$400 million, it could trigger a 5-7% correction in TRI shares. However, the base scenario is that Bloomberg will cede the US legal AI market, focusing on Europe and Asia, where Thomson Reuters has weak positions.
Editorial Forecast
Asset: Thomson Reuters shares (TRI on NYSE) — moderate growth in the next 24–72 hours.
Key Levels: Current level — $162. Nearest resistance — $165 (2026 high), support — $158. If $165 breaks, next target is $170.
Confidence Level: High. The deal is already approved, no negative surprises expected, analysts are upgrading ratings.
Main Risk: A sudden announcement from the Department of Justice about expanding the antitrust investigation — although unlikely since the deal is already agreed, any mention of the DOJ in the news could cause short-term panic.
This analysis represents the editorial opinion and is not individual investment advice.
— Editorial Team
