Software ate the world. Now AI is eating software, and in the world of legal technology, Thomson Reuters is trying to stay off the menu.
The company runs one of the largest hoards of legal data on the planet. It’s combining this information trove with generative AI to offer new products, such as Westlaw Advantage and CoCounsel, that help automate legal work.
The strategy was a hit with investors earlier this year, but since the summer, new rivals have grown bigger, bringing questions about Thomson Reuters’ AI future.
A potentially more troubling question: Can CoCounsel produce legal intelligence that’s better than what ChatGPT and existing big AI models can already provide? In essence, is Thomson Reuters working on a “model wrapper,” or something with more enduring value?
The company’s edge
The company made its pitch on Tuesday when it reported quarterly results, and the response from Wall Street was tepid at best. Thomson Reuters shares fell more than 6%, leaving them down more than 30% since mid-July.
The Toronto-based company said its legal unit grew organic revenue 9% to $700 million, up from 8% growth in the first half of the year. AI features drove “double-digit growth” across its CoCounsel products.
On a call, Thomson Reuters CEO Steve Hasker told analysts that virtual “agents” are taking on more of a lawyer’s work. The company’s edge, he argued, is the one-two punch of its law library and editorial content.
“Difficult, if not impossible, to replicate,” he added.
Related stories
Business Insider tells the innovative stories you want to know
Business Insider tells the innovative stories you want to know
New challengers
These days, Thomson Reuters is playing defense as much as offense in this lucrative part of the tech industry. While the company has been a heavyweight in legal tech for decades, the rise of generative AI cracked the door for challengers, and a wave of new entrants has rushed in as venture funding piles up.
In June, RELX’s LexisNexis unit formed a strategic alliance with AI legal tech startup Harvey. The partnership brought together the world’s other huge corpus of legal data with a technical team capable of applying generative AI to this information trove. Now, when lawyers fire up Harvey’s service, its outputs include content and citations from LexisNexis, although a LexisNexis login is still required.
In late August, another challenger emerged when legal operations platform Clio acquired vLex, a specialist in legal data, for $1 billion.
Thomson Reuters’s Hasker dismissed frothy AI startup valuations and bold growth claims, calling them “squishy,” during the call on Tuesday. He also said CoCounsel adoption is outpacing rivals.
A CoCounsel moat
The CEO framed Thomson Reuters’ moat as a powerful combination of its Westlaw legal data and human-in-the-loop editing and validation, arguing this is essential in an industry where accuracy, timeliness, and security are paramount.
“Litigation is high-stakes work with no room for error and significant consequences for being wrong,” Hasker said.
Its Westlaw product ingests more than 300 million documents a year from hundreds of court systems and tens of millions of rulings. An in-house bench of attorneys and editors turns this scramble of data into court-safe guidance, writing headnotes, verifying facts, and tagging issues to make the content easier to find. Roughly 85% of primary documents in Westlaw, such as cases and statutes, carry these editorial upgrades.
The strategy now is to push beyond retrieval and summarization and into more generative services that help lawyers with more valuable tasks. That includes agentic workflows.
The OpenAI question
The question hanging over everyone in legal tech, Thomson Reuters included, is whether OpenAI jumps in. This AI giant could sell a vertically packaged app to law firms and in-house teams, though there are no clear signs this is in the cards right now. Still, fears flared recently after the startup posted demos using its own models and in-house AI software for contract review, a staple of the legal tech sector.
If OpenAI goes there, it could risk cannibalizing customers who already pay to build on its models. Indeed, CoCounsel uses large language models including OpenAI’s GPT offerings. (OpenAI is also an early investor in Harvey, which only sharpens the tension here).
Even without a bespoke legal product, OpenAI is already the gravitational competitor. Lawyers already use ChatGPT, either officially or otherwise, and other business people are increasingly tapping the chatbot for legal input.
Harvey’s CEO has said buyers benchmark its legal AI solutions against the latest general-purpose models. That comparison means OpenAI can shape the market by shipping upgrades, keeping the rest of the field on alert.
Hasker’s counterargument is that depth wins. Specialized AI tools that tap unique and accurate legal data could gain ground against more general-purpose AI offerings.
“Customers are starting to understand the difference,” he said, before adding an honest forecast. “Anyone who will tell you they know exactly what’s going to happen in this environment is probably slightly deluded.”
Have a tip? Contact this reporter via email at mrussell@businessinsider.com or Signal at @MeliaRussell.01. Use a personal email address and a non-work device; here’s our guide to sharing information securely.
