Thomson Reuters is getting firmer proof that lawyers are actually using artificial intelligence day to day, but the rapid pace of new product launches is also raising the risk that pricing gets tougher. A note from RBC Capital Markets, a Canadian bank, argues that the swing question for the information and analytics giant is no longer whether legal AI has staying power, but how big the market can get.
One data point caught RBC's eye: Harvey, a legal AI startup, reported that monthly “token” use—a measure of AI workload—jumped 12 times from January to May. That kind of leap suggests these tools are moving from trial phases into regular workflows, a key signal for investors watching whether the technology will generate sustainable subscription revenue.
What the Data Shows
Tokens are the basic units of input and output that AI models process. A 12x increase in token usage over four months implies that law firms and legal departments are not just testing Harvey but integrating it into daily tasks like document review, contract analysis, and legal research. For Thomson Reuters, which owns the Westlaw legal research platform and the CoCounsel AI assistant, this is a double-edged signal.
On one hand, it confirms that demand for legal AI is real and growing. On the other, it shows that competitors like Harvey are gaining traction, which could pressure Thomson Reuters to keep its own offerings competitive on both features and price. The company recently gave CoCounsel Legal customers access to CoCounsel Next, an upgraded version of its AI assistant, while Anthropic—an AI lab backed by major tech investors—launched Claude Tag, a tool that lets users tag and organize legal documents. These rapid product releases highlight how quickly features are spreading across the category.
Why It Matters for Investors
For everyday investors, the key takeaway is that the legal AI market is moving from hype to reality, but the economics are still being worked out. Usage-based metrics like tokens make it easier for buyers to compare tools on cost per document, query, or matter. If customers start shopping based on “price per output,” pressure may show up first as slower revenue growth per user, not immediate cancellations.
RBC maintained an outperform rating on Thomson Reuters stock and a price target of $129, even after the shares rose 3.8% on Friday to $84.11. That target implies significant upside, but it depends on the company's ability to scale revenue without letting unit prices get competed down. The broader backdrop is also relevant: demand for AI across sectors has been a major theme in markets, as seen in recent rallies in chip stocks and price hikes in memory chips driven by AI demand. But the legal industry has its own dynamics, including conservative buying cycles and high sensitivity to cost.
Thomson Reuters is not the only company navigating this shift. In other sectors, companies like H&M are facing fading margin gains as demand softens, while South Korea's growth hinges on chip exports. The common thread is that demand signals are becoming more nuanced, and investors need to look beyond top-line growth to understand pricing power.
What to Watch Next
For Thomson Reuters, the debate shifts from “is demand real?” to “can revenue scale without unit prices getting competed down?” That is a key driver of how much valuation support targets like $129 can command. Investors should watch for quarterly updates on CoCounsel adoption rates, average revenue per user, and any commentary on pricing trends. The company's next earnings report will likely provide more clarity on whether the 12x token jump at Harvey is a leading indicator for the whole market or a sign that competition is fragmenting the opportunity.
In the meantime, the rapid product launches from both Thomson Reuters and rivals like Anthropic suggest that legal AI is becoming a standard tool rather than a novelty. For everyday investors, that means the opportunity is real, but the path to profitable growth may be bumpier than early hype suggested.
