Thomson Reuters announced several new generative artificial intelligence tools and platforms earlier this week. I had the privilege of attending a press briefing hosted by Thomson Reuters in advance of the announcement. Thomson Reuters personnel discussed the new products and offerings and allowed several of us to try them out. The presentations and products were impressive.
Several articles review these products in detail, and I won’t attempt to duplicate those. Suffice it to say, the main offerings discussed were:
- An AI assisted research tool on Westlaw Precision
- CoCounsel integrations including a GenAi assistant and CoCounsel Core that move from keyword based inquiries to natural language inquiries (to be launched in the future)
- A quotation analysis (to be launched in the future)
- A more robust Practical Lawyer tool
That is all well and good, but there were several more general and attitudinal takeaways from my time with Thomson Reuters representatives and the press briefing that may be more important.
First, several of us wondered what might happen to Casetext, the creator of the original GenAI tool for legal, CoCounsel, after its acquisition by Thomson Reuters. Would Casetext’s scrappiness and entrepreneurial sprint get lost in Thomson Reuters’ bureaucracy? Would Thomson Reuters overrun the Casetext culture? Would Casetext’s innovative approach to legal technology wither away?
The Casetext CoCounsel product is playing a central and dynamic role in the Thomson Reuters suite of offerings.
I’m happy to say based on what I heard at the briefing, it looks like the Casetext CoCounsel product is playing a central and dynamic role in the Thomson Reuters suite of offerings. Rather than withering away, there seems to be a symbiotic relationship between the companies and their personnel. This relationship can only help both companies and, more importantly, their users.
The attitude of Thomson Reuters representatives I talked to seemed to be one of admiration for Casetext. They had a real appreciation of Casetext’s accomplishments and knowledge. Thomson Reuters personnel readily admitted that Casetext had a head start in developing GenAI legal products on which it wants to capitalize. On the Casetext side, Jake Heller, the Casetext co-founder and now head of CoCounsel at Thomson Reuters, was a key presenter at the briefing. Heller told me that the opportunity to test offerings through the Thomson Reuters system offered tremendous opportunities that the small Casetext organization could not match.
Even just watching the interaction between Heller and the Thomson Reuters representatives, I sensed an excitement and energy about working together. The Casetext CoCounsel product is featured in the Thomson Reuters dashboards and is being promptly integrated into the existing Thomson Reuters products. All this bodes well for Thomson Reuters and future offerings and innovations.
Thomson Reuters offers a consistent and clear vision as to where Thomson Reuters wants to go with GenAI
Second, the Thomson Reuters presenters offered a consistent and clear vision at the briefing as to where Thomson Reuters wants to go with GenAI. The moon shot goal, they say, is to develop an AI assistant or agent for every lawyer and legal professional within five years. This assistant will handle those tasks that need to be done but that a lawyer or legal professional doesn’t need to spend time doing. I have heard this assistant concept articulated by some in the nonlegal tech community (Leo Laporte recently discussed this concept at some length in his recent Ask the Tech Guys podcast, for example). But I have not heard this clear vision from other AI providers in the legal space. At least not as clearly as Thomson Reuters.
As David Wong, Thomson Reuters Chief Product Officer put it, Thomson Reuter’s vision is to provide within five years an AI assistant that speaks the language of lawyers. To which lawyers can delegate professional tasks. That can do the delegated work at super speed and accuracy. An assistant to whom professional tasks can be reliably delegated. An assistant who can do this work at super speed and accuracy.
This vision reflects another attitude that came out at the briefing. In developing its AI products, Thomson Reuters personnel confirmed that they focused not on what AI can do but on the jobs that need to be done that AI could do, at least as a starting point. The presentation seemed to reflect this attitude in the products discussed. I have seen too many vendors in the AI space try to rush to market products just to claim they have a product without regard to whether that product solves problems and relieves a pain point legal professionals’ experience.
It seemed important to Thomson Reuters that the people using their tools could easily master and understand how to use them
Another takeaway from the briefing. It seemed important to Thomson Reuters representatives at the briefing that the people using their tools could easily master and understand how to use them. Which, in turn, reflects the attitude that AI tools should be useful to the specialized market of lawyers and legal professionals.
Thomson Reuters representatives were careful to point out, for example, that they were not enamored with the notion that lawyers and legal professionals need prompt engineering training to use Thomson Reuters products. They don’t encourage users to focus on trying to write prompts instead of just asking questions in language that lawyers and legal professionals use. The representatives did say that there may be instances where prompt engineering tools might be necessary. But, by and large, Thomson Reuters has made a conscious effort to embed the prompt engineer concepts into the product. As a result, the user doesn’t have to master those concepts or be burdened by them.
What stood out from Thomson Reuters’ new product announcements was the attitude behind them
As Wong put it, creating the prompt engineering to use the product was Thomson Reuter’s job rather than the user’s. This approach reflects a basic understanding of law firm culture. Every minute most lawyers spend learning prompt engineering is money out of pocket since they can’t (or shouldn’t) bill for it.
What stood out from Thomson Reuters’ new product announcements was the attitude behind them. The spirited integration with Casetext. The clear vision is to create a ubiquitous virtual assistant. And the recognition that ease of use is so crucial to the users of legal technology products. While I can’t vouch for the new products today offered or what will come down the pike, I can vouch for the attitude and the culture the attitudes suggest. And that is promising for the future.