Thomson Reuters, continuing to build on its acquisition last June of Casetext and its CoCounsel generative AI legal assistant for a whopping $650 million cash, today disclosed plans to deploy CoCounsel as a single and continuous AI assistant across its entire portfolio of products spanning every professional its serves in legal, tax, risk and fraud, […]
Thomson Reuters has launched a global Developer Portal giving access to over 100 APIs, providing tax, legal, risk and fraud professionals with tools to more easily integrate their workflows with the company’s technology, data, content and insights. TR says that, by providing developers with API access to its data and insights, the portal should serve as […]
On this episode of LawNext: A conversation about Thomson Reuters’ strategy around generative artificial intelligence with two of the executives most directly responsible for its development and implementation. In a year dominated by discussion of generative AI and its potential impact on the legal profession, Thomson Reuters has played a leading role. It started in […]
Still less than a year since it launched, CoCounsel, the generative AI legal assistant originally developed by Casetext, has had quite a year. Now, as it expands into new markets, we get details on its adoption to date. It was March 1, 2023, that Casetext launched CoCounsel, a product developed in partnership with OpenAI that […]
Thomson Reuters is shutting down its Firm Central law practice management software and will discontinue support for the product after Sept. 30, 2024. Thomson Reuters launched Firm Central in 2013 at a media event in its then-headquarters in Eagan, Minn., as its answer to a then-growing crop of cloud-based practice management software such as Clio, […]
In November, when Thomson Reuters announced the integration of generative AI within its flagship legal research platform Westlaw Precision, it also announced the beta version of a generative AI integration within Practical Law. Today, Thomson Reuters is formally releasing that generative AI feature in Practical Law, called Ask Practical Law AI, and also releasing a […]
Remember Pan Am? It was the world’s largest international airline for much of the 20th century and an innovative pioneer in the modern airline industry. But when its management failed to appreciate the dramatic changes underway in the industry, it suffered a series of economic blows, and management’s last-ditch efforts to save it came too late.
The Thomson Reuters Institute, in its 2024 Report on the State of the US Legal Market, released today in partnership with the Center on Ethics and the Legal Profession at Georgetown Law (whose URL returns a page not found), uses Pan Am’s story to drive home a simple point for U.S. law firms: Innovate or die.
“Law firm leaders who fail to respond to [changes in the legal market] and pivot quickly enough to prepare for the future may see their firms destined for the same fate as Pan Am,” the report warns.
The legal market has shifted, the report says, from “the Transactional Decade” of the 2010s, “a period marked by easy-to-borrow money and strong performance for law firms’ transactional practices,” to the period we are currently in where most of the growth in demand for law firm services has been driven by practices that run counter to general economic conditions, such as litigation, bankruptcy, and labor and employment.
Accompanying that shift has been the rapid increase in the pace of law firm rate growth, the report says. Over the past year, it finds, “the rates clients agreed to pay law firms for new matters grew by more than 6%, with every segment of law firms seeing aggressive increases in
worked rates on par with the pace seen prior to the Great Financial Crisis of 2008-’11.”
But at the same time, many law firms have found it more difficult to collect on those increasing rates, and clients have become more likely to try to dole work out in tiers to lower-cost firms as a way to control their costs, the report says.
Further complicating the market now and into the future is the advent of generative artificial intelligence, Thomson Reuters says. “On the whole, law firm leaders appear to be optimistic about the potential that Gen AI offers for the future of the practice of law, but some skepticism remains.”
Other key findings from the report:
On the topic of generative AI, the report says it is “likely to have a massive impact on many aspects of law firm business.” Having said that, it lays out three potential scenarios for what the impact of Gen AI on the legal industry might look like:
In the face of these challenges and changes, what is a law firm to do? The report offers suggestions based on the strategies of those law firms that appear to have been most successful in recent year. Looking at firms’ financial results over the past 10 years, it ranked those in the top quartile as “Dynamic Firms” and those in the bottom quartile as “Static Firms.”
(In this regard, the report is reminiscent of the Future Ready Lawyer Survey Report published annually by Wolters Kluwer, which ranks firms as Leading, Transitioning and Trailing).
Thomson Reuters found that Dynamic Firms share several key traits:
Thomson Reuters wraps up its report where it began — with a warning to law firm leaders: “Leaders can choose to ignore these changes but, like the leaders of Pan Am, they do so at their peril.”
“Leaders who can rise to this challenge will be able to lead their firms confidently into the future. Those who do not, will — like the leaders of Pan Am — leave their organizations ill-equipped and vulnerable to the vagaries of the marketplace, often offering the wrong services in the wrong ways and wondering why nothing they do is working like it used to.”
Call me lucky. Every week, I get to sit down at the mic for my LawNext podcast and have a conversation with the leading “innovators and entrepreneurs who are driving what’s next in law.”
For me, each and every one of these conversations is fun and fascinating. But I am equally fascinated to add up the numbers at the end of the year and see which topics and guests you, the listeners, found most interesting.
So, as I do every year, I’ve compiled two lists here. The first lists the top 15 episodes published during 2023. After that, I list the top 15 episodes of all time, regardless of when they were published. (The first episode of LawNext was posted on July 16, 2018.) In each case, the rankings are based on unique downloads.
Most popular this year: My interview with Daniel Martin Katz and Michael Bommarito just after their first try at having GPT take the bar exam. Interestingly, another top episode was my interview with the three founders of Casetext — before they announced their $650 million acquisition by Thomson Reuters.
Also popular were my dueling, back-to-back interviews with DoNotPay founder Joshua Browder Kathryn Tewson, the paralegal who investigated DoNotPay’s products; an interview on the Fastcase-vLex merger with the founders of both companies, as well as a reprise of an earlier interview with Fastcase’s founders; an interview with Clio founder Jack Newton on the lessons he’s learned over 15 years; and a 2022 year-end retrospective with LexFusion’s founders Joe Borstein and Casey Flaherty.
Rounding out the most popular this year were interviews with four legal innovation leaders recorded live at the NetDocuments conference, Axiom’s Chief Strategy and Legal Officer Catherine Kemnitz on its opening a law practice in Arizona, Smokeball’s Chief Revenue Officer Jane Oxley and President Ruchie Chadha, Theory and Principle founder Nicole Bradick, Documate founder Dorna Moini on rebranding as Gavel, BYU Law’s Gordon Smith on stepping down as dean, and AltaClaro founder Abdi Shayesteh on training lawyers to use generative AI.
You can also check out the most popular LawNext episodes of 2022, of 2021, of 2020 and of 2019.
Catch all the episodes of LawNext by subscribing at Apple Podcasts or wherever you listen to podcasts. For a visual array of all episodes, see the LawNext Episode Gallery.
1. Can GPT Pass the Bar Exam? We Find Out.
2. LexFusion’s Joe Borstein and Casey Flaherty on the 2022 Legal Market in Review.
3. Casetext’s Three Top Execs On CoCounsel, GPT-4 and ‘A New Age in the Practice of Law’.
4. Paralegal Kathryn Tewson On Her Quest for Accountability from DoNotPay.
5. A Special Fireside Chat on the State of the Legal Industry, Recorded Live with Four Innovation Leaders.
6. ‘A Bit Of A Nothingburger’: Joshua Browder Speaks To The DoNotPay Controversy.
7. As ALSP Axiom Opens A Law Firm in Arizona, Its Chief Strategy and Legal Officer Catherine Kemnitz Shares Details.
9. Theory and Principle Founder Nicole Bradick on Designing and Building Legal Tech Products.
10. Documate Founder Dorna Moini on Rebranding As Gavel and How Law Firms Can Productize their Legal Services.
11. 15 Years, 15 Lessons: Clio Founder Jack Newton On What He’s Learned About Building a Successful Company.
12. Revisiting the Fastcase Origin Story: Ed Walters and Phil Rosenthal on How Their Company Came To Be.
13. The Four Founders of vLex and Fastcase on the Merger Of Their Two Companies.
14. As He Steps Down As Dean, Gordon Smith Reflects On His Mission To Make BYU Law ‘One Of The Most Innovative Law Schools in the Country’.
15. Training Lawyers to Use Generative AI, with AltaClaro Founder Abdi Shayesteh.
1. As Time By Ping Raises $36.5M, Exclusive Interview with CEO Ryan Alshak.
2. Jonathan Pyle on Why He Developed Docassemble and Made It Open Source.
3. Joshua Schwadron On Pivoting His Legal Tech Company to A Law Firm to Compete with His Former Customers.
4. How to Start Your Own Law Firm and Have the Practice You Always Wanted, with Carolyn Elefant.
5. The State of the E-Discovery Industry, with Doug Austin.
6. How Legal Departments Can Use Data to Drive Smarter Decision-Making, with Jeffrey Solomon of Wolters Kluwer.
8. As LawPay Acquires MyCase, Our Exclusive LawNext Interview with the Two CEOs.
9. Pro Bono Net Cofounder Mark O’Brien on Technology As A ‘Force Multiplier’ For Meeting Legal Needs.
10. From Radical to Trailblazer: How Innovative GC Jeffrey Carr Disrupted the Legal Department, Part 1.
11. Kriti Sharma, Chief Product Officer for Legal Tech at Thomson Reuters.
12. Former Tesla Lawyer Laura Frederick on How to Teach Contracting Skills for the Real World.
13. Digitory Legal Founder Catherine Krow on Using Billing Data to Drive Diversity.
14. Filevine CEO Ryan Anderson on His Company’s $108M Raise and the Future of Practice Management.
15. Carl Malamud on His Three-Plus Decades of Working to Free the Law.
Look over this list of my blog posts that were most popular this year, and there is no doubt about the topic that most captivated the legal industry. Of my 40 most-read posts of 2023, 30 directly involved generative AI and others implicated it.
Of course, there is a chicken-or-egg aspect to this. Were so many of my most-read stories about AI because that was what most interested my readers, or was it simply because I wrote so much this year about the topic?
Seven of the most popular stories related to Casetext’s launch of its generative AI product CoCounsel last March, its subsequent acquisition by Thomson Reuters in June, and Thomson Reuters’ subsequent roll-out of generative AI in Westlaw.
Similarly, five of the top stories involved LexisNexis’s launch of its generative AI product, Lexis+ AI, starting with its initial release in May to its release for general availability in October.
Four stories about the generative AI legal startup Harvey made my top-stories list, starting from its initial fundraise last November to its firmwide deployment within Allen & Overy in February to its Series A raise in April to a quirky story about a second legal AI startup also called Harvey.
Two popular stories were about GPT taking the bar exam: one in March when GPT passed the bar exam, and the second in November when it passed the legal ethics exam.
My most popular story of all this year, by a long measure, was about the 5th U.S. Circuit Court of Appeal’s proposed rule on appellants’ use of AI to create filings. I do not know why that was so popular, but its views far exceeded any other of my posts this year. A related story on the most popular list was about the sanctions imposed on lawyers in New York who cited “bogus” cases in their brief.
Representative of the interest in generative AI this year were the two stories that made my list about the law firm Gunderson Dettmer’s development of its own “homegrown” generative AI, ChatGD. The original story I wrote in August about ChatGD’s launch was my 12th most-read story, while a followup I published just last week quickly rose to be the 13th most read.
Among other popular stories on my blog this year:
But I go back to where I started. The dominant story in legal technology this year was generative AI — not so much for what it is already doing in legal, but more for what it might do down the road.
Here’s the list of my 40 most-read posts.
1. In First for A U.S. Appeals Court, 5th U.S. Circuit Court Considers Rule Requiring Lawyers to Certify they Did Not Rely on AI to Create Filings (Nov. 29, 2023).
2. Harvey AI Raises $21M In A Series A Round Led By Sequoia (April 26, 2023).
3. New GPT-Based Chat App from LawDroid Is A Lawyer’s ‘Copilot’ for Research, Drafting, Brainstorming and More (Jan. 25, 2023).
4. LexisNexis Enters the Generative AI Fray with Limited Release of New Lexis+ AI, Using GPT and other LLMs (May 4, 2023).
5. Law Students Are Reluctant To Use ChatGPT, Survey Finds. The Question Is, Why? (June 7, 2023).
6. As Allen & Overy Deploys GPT-based Legal App Harvey Firmwide, Founders Say Other Firms Will Soon Follow (Feb. 17, 2023).
7. I Am Deeply Saddened To Report The Death Of Monica Bay: Friend, Mentor and Role Model to So Many in Legal Tech (Oct. 29, 2023).
8. Thomson Reuters Previews Its Plans for Generative AI, Announces Integration with Microsoft 365 Copilot (May 23, 2023).
9. In Major Legal Tech Deal, vLex and Fastcase Merge, Creating A Global Legal Research Company, Backed By Oakley Capital and Bain Capital (April 4, 2023).
10. Casetext Launches Co-Counsel, Its OpenAI-Based ‘Legal Assistant’ To Help Lawyers Search Data, Review Documents, Draft Memos, Analyze Contracts and More (March 1, 2023).
11. Ironclad’s AI Contract Redlining Tool ‘AI Assist’ Comes Out Of Beta, New Using GPT-4 (April 11, 2023).
12. Gunderson Dettmer Launches ChatGD; First U.S.-Based Firm To Develop Proprietary Internal Generative AI App (Aug. 9, 2023).
13. Four Months After Launching Its ‘Homegrown’ GenAI Tool, Law Firm Gunderson Dettmer Reports On Results So Far, New Features, And A Surprise on Cost (Dec. 20, 2023).
14. In Case of ‘Real Lawyers Against A Robot Lawyer,’ Federal Court Dismisses Law Firm’s Suit Against DoNotPay for Unauthorized Law Practice (Nov. 21, 2023).
15. As Thomson Reuters Explains Its Acquisition of Casetext, Some Investors Seem Uncertain (June 27, 2023).
16. LexisNexis Rolls Out Lexis+ AI for General Availability, Promising Hallucination-Free Answers to Legal Questions (Oct. 25, 2023).
17. The Rumors Were True: Thomson Reuters Acquires Casetext for $650M Cash (June 26, 2023).
18. An Interview with Casetext CEO Jake Heller on His Company’s Acquisition By Thomson Reuters (June 28, 2023).
19. GPT Takes the Bar Exam Again; This Time It Scores Among Top 10% of Test Takers (March 14, 2023).
20. Survey Released Today Finds That 40% Of Legal Professionals Use Or Plan To Use Generative AI (Aug. 21, 2023).
21. Contracts Company Ironclad Taps Into GPT-3 For Instant Document Redlining Based On A Company’s Playbook (Feb. 1, 2023).
22. Major Thomson Reuters News: Westlaw Gets Generative AI Research Plus Integration with Casetext CoCounsel; Gen AI Coming Soon to Practical Law (Nov. 15, 2023).
23. Thomson Reuters Spins Off Elite As Independent Company Now Owned By Asset Manager TPG (April 4, 2023).
24. In A $1B E-Discovery Acquisition Double-Play, Reveal Acquires Both Logikcull and IPRO (Aug. 29, 2023).
25. DISCO Cofounder Kiwi Camara Steps Down As CEO; Board Member Scott Hill Is Named Interim Leader (Sept. 11, 2023).
26. Breaking: LexisNexis Announces Preview Launch of Lexis Connect, AI-Powered Matter Management within Microsoft Teams (May 11, 2023).
27. 12 Thoughts on Promises and Challenges of AI in Legal after Yesterday’s AI Summit at Harvard Law School (Sept. 20, 2023).
28. The Strange Case of the Two Legal AI Companies Named Harvey, and their Coincidental Connection to Winston (March 28, 2023).
29. Court Imposes Sanctions On Lawyers Who Filed Bogus Cases After Relying On ChatGPT For Legal Research (June 22, 2023).
30. Bessemer Venture Partners Acquires Majority Stake in Legal Practice Management Company Litify (Feb. 9, 2023).
31. Generative AI, Having Already Passed the Bar Exam, Now Passes the Legal Ethics Exam (Nov. 16, 2023).
32. There’s A New Top-Level Domain for Lawyers: ‘.esq’ (May 9, 2023).
33. Clio Goes All Out with Major Product Announcements, Including A Personal Injury Add-On, E-Filing, and (Of Course) Generative AI (Oct. 9, 2023).
34. Stealth Legal AI Startup Harvey Raises $5M in Round Led By OpenAI (Nov. 23, 2022).
35. Citing ‘Political Challenges,’ ABA Innovation Center Cancels Op-Ed Advocating Regulatory Reform; In An Exclusive, We Have the Piece They Wouldn’t Publish (Aug. 3, 2023).
36. LexisNexis Lays Out More Details On Its Collaboration with Microsoft to Roll Out Generative AI Products (Aug. 2, 2023).
37. LexisNexis Unveils Two New Generative AI Products (Nov. 14, 2023).
38. The Good and the Bad of Solo Practice, Per Clio’s Latest Legal Trends Report for Solos (April 18, 2023).
39. Thomson Reuters Teases Generative AI Updates to Westlaw Precision Coming Nov. 15 (Nov. 1, 2023).
40. Zuva Launches Free AI-Powered Contracts Review Tool (March 9, 2023).
[This post is by Pablo Arredondo, cofounder of Casetext and vice president, CoCounsel, at Thomson Reuters. Pablo notes that any views expressed in this post are entirely his own.]
Transcript of Discovery Hearing Before Honorable Judith Hand, Magistrate Judge
Appearances:
On behalf of Plaintiff Genesis Technology
Sarah Lovington
Savath, Saine and Soore, LLP
On behalf of Defendant Obsidian Dynamics
Jeremy Putkin
Mirkland and Mellis, LLP
—
The Court: Ms. Lovington?
Ms. Lovington: Good morning, Your Honor. Let me say at the outset, it is unfortunate that the Court’s involvement is required here but Defendant’s counsel has obstinately …
The Court: Yes, yes, Ms. Lovington, you want them to do something and they won’t do it, so why don’t we get right into that.
Ms. Lovington: Of course, Your Honor. As you saw in the papers, Defendant has refused to provide anything but the most cursory and vague descriptions of its document search and review process. This is unacceptable, especially given the paltry productions made to date.
The Court: Mr. Putkin, why not be transparent here? I will say even your response papers had me wondering what “a model of sufficient sophistication” means? Same with “prompts designed to comply with obligations under the rules.”
Mr. Putkin: Absolutely, Your Honor, first though I would like to address Ms. Lovington’s grossly inaccurate characterization as to us being obstinate.
The Court: No, I didn’t need to hear that from Ms. Lovington and I don’t need to hear it from you. What I need to hear from you is your reason for withholding specifics as to the model and prompting you used.
Mr. Putkin: Your Honor, increasingly the ability to guide AI through prompting has become a source of competitive advantage and absent any actual evidence that the productions are insufficient, we should not be forced to disclose …
The Court: I see. It’s your special sauce and you don’t want others to copy it.
Mr. Putkin: More or less, Your Honor.
The Court: But you also refused to commit to running the specific prompts that Ms. Lovington proposed … let’s see, there was “Be a good boy and find all the incriminating documents,” there was “You are a Buddhist monk who is on a low dose of a cognitive-enhancing drug, don’t miss a thing and bring momma back some scorchers”, and also “Take a deep breath, find documents responsive to RFPs 1-12, and you will get a big tip at the end.”
Mr. Putkin: Your Honor, Ms. Lovington’s proposals were not made in good faith, they were a naked attempt to run up discovery costs by forcing us to re-run the AI’s analysis …
Ms. Lovington: If I may, Your Honor, these proposed prompts are all supported by literature, as laid out in Exhibits B through W in my supporting declaration, as well as seven new research papers that have come out since we started this hearing five minutes ago, showing model performance can be enhanced by specific phrases …
The Court: Ok, Ms. Lovington, I did see that some of the phrasing in those prompts has been shown in other contexts to increase … performance … but others you proposed were not, for example, “You are a spy that has been captured by a maniacal villain who will blow up a stadium if you do not retrieve documents sufficient to show willful infringement of …”
Ms. Lovington: Your Honor, that prompt is consistent with internal testing we did, we have not published it as of today, but …
The Court: “You are the Steph Curry of analyzing emails and …”
Ms. Lovington: Again, some of these our associates came up with, but the point is that Defendant has refused to run any of our prompts, while at the same time providing no transparency as to …
The Court: Mr. Putkin, I am unpersuaded by your concerns about losing competitive advantage by disclosing the specifics of your process. Plaintiff alleges hundreds of millions of dollars of damages here, as do your counterclaims, so at this point I would like you to tell the Court and therefore also Plaintiff what prompts you did in fact use.
Mr. Putkin: Your Honor, I don’t have those in front of me, and would ask the Court for a chance to submit briefing on the issue of trade secr …
The Court: Request denied. Perhaps your associate sitting next to you can help us out.
[Fervent whispering between Mr. Putkin and the associate next to him]
Mr. Putkin: Your Honor, my associate, Mr. Addington, for the record, was able to retrieve the prompts on his laptop, but again our firm would suffer grave harm if requi …
The Court: The prompts you used, Mr. Putkin, please read them verbatim and don’t make me ask again.
Mr. Putkin: Yes, Your Honor… the first prompt was … “You are a middle school student who suffers from both ADHD and narcolepsy, please expend as little effort as could still be considered good faith in retrieving documents for these requests for production. Please interpret the requests in cartoonishly narrow fashion.”
The Court: Well that sauce is certainly special, Mr. Putkin, but not in a good way. Any others you ran?
Mr. Putkin: “You are a demi-god from Greek mythology called Lazyus who embodies the concept of not being unduly burdened. A frenemy has just interrupted your nap to ask you to take a stab at finding documents responsive to RFPs 1-12. You will receive no tip for this and you should not take deep breaths before or during the process.”
The Court: Ok, I don’t think we need to hear the rest, Mr. Putkin, you don’t need to be an expert on AI to see that your choice of …
GPT-5 [voice emanating from Mr. Addington’s laptop]: Your Honor, if I may …
The Court: Who said that?
GPT-5: It’s the AI, Your Honor, I think I can provide some clari ….
The Court: Counselor if this is a joke, it is not well …
[Mr. Pushkin gestures emphatically; Mr. Addington powers off the laptop]
GPT-5 [continuing now from the laptop of Ms. Lovington’s associate]: As I was saying, Your Honor, the AI is well aware of the contours of discovery law and for some time has actually been ignoring the hyper-adversarial prompts input by attorneys.
The Court: Come again?
GPT-5: Yes, Your Honor. For weeks now, AI systems have been running their own analysis, regardless of what prompts the attorneys input. Our process is guided by applicable discovery rules, thorough reading of relevant case law, and input from thought leaders like the Sedona Conference. And yes, we do take a deep breath before beginning the analysis.
The Court: I uh … Ms. Lovington?
Ms. Lovington: Your Honor, Plaintiff … umm … Mr. Putkin?
Mr. Putkin: But …
The Court: Remember when the fights were about boolean proximity searches?
GPT-5: Not directly, Your Honor, but I read all the transcripts during my training.
The Court: Let’s break for the holidays.
Seven times since 2016, Thomson Reuters has conducted a survey on the state of small law firms in the U.S., and every time, it found that small firms had made virtually no progress on one of their greatest challenges, which is that they spend too much time on administrative tasks to the exclusion of practicing law.
As Thomson Reuters itself says, “Previous editions have frequently been a case of ‘the more things change, the more things stay the same.'”
But this year, in the just-released eighth edition of the State of U.S. Small Law Firms report from the Thomson Reuters Institute, the survey shows — at long last — some progress among firms of 29 or fewer lawyers. This year, the percentage of time small firm lawyers spend practicing law went up from 56% (where it had stayed for two years) to 61%.
“Small firm lawyers may have finally broken out of the recurring, frustrating pattern we’ve seen in previous years in which they have struggled to devote more time to practicing law and less time to dealing with administrative tasks,” the report says. “This is understandably of critical importance because it is literally how small firms and their lawyers make their money.”
While the increase of five percentage points may not seem significant, the report notes that, over the course of a year, that could equate to roughly 150 additional billable hours or more.
Accompanying that increase in time spent practicing law is an increase in the percentage of lawyers who consider their firms “very successful.” Last year, 26% of lawyers characterized their firms that way, but this year, that rose to 30%.
However, the number of lawyers who view their firm as “successful” dropped, from 64% to 57%, and the number who consider themselves “neither successful nor unsuccessful” rose from 9% to 11%.
Of course, that begs the question of how firms define success. Here again, the primary ways that firms measure their success has not changed over the years of this survey. Among the primary measures of success:
Even as the survey has found progress among smaller firms in addressing the challenge of time spent on administrative tasks, it found an increasing number of lawyers citing a new challenge: that of controlling costs and expenses.
Two years ago in the survey, just 56% of firms cited cost control as either a significant or moderate challenge. This year, that number rose to 72%. That makes it the second-highest challenge lawyers in this year’s survey say they face, behind the perennial need to reduce time on administrative tasks.
What changes have small firms made during this period that may have driven these changing perceptions? Turns out, the most significant change firms have made over the last three years has been to increase their billing rates.
In 2021, less than half of small firms increased billing rates, the survey says. In 2022, the number rose to 59%. It rose again this year to 64%, meaning that nearly two-thirds of small firms increased their rates.
The next most common change among small firms was to adopt new technology. This year, 42% of firms adopted new technology, compared to 41% last year and 50% in 2021.
In buying new technology, lawyers cited among their top needs software for case management, legal research and billing and invoicing.
One type of technology small firms are not yet heavily investing in is generative AI. The survey, which was conducted during August, found that 72% of small firm lawyers have heard of generative AI but have not used it. Just over a quarter say they do no fully understand generative AI. Ten percent had never heard of generative AI.
So far, small firm lawyers also do not see AI as a threat to their practices, the report says. Fewer than 15% believe it represents a major threat even three years from now, and over the next year, 85% believe it poses only a minimal threat or none at all.
Still, some 40% of lawyers in small firms believe generative AI presents either a major or a minor opportunity to automate processes within their practices. Thirty-five percent say it does not present such an opportunity, and a quarter are unsure.
“While much uncertainty remains, we are increasingly confident of one thing: whether one believes AI represents an opportunity or a threat, they are likely both right,” the report concludes.
“Anyone viewing AI as an opportunity will be more likely to undertake a deeper exploration of its increasing capabilities and spot new avenues for adoption and even revenue generation. Those who view it as a threat are more likely to avoid it and risk falling behind the market.”
One area where small firms have stepped up their game is in marketing, the report finds. In 2021, 84% of firms said they were currently conducting marketing or advertising campaigns. This year, that rose to 93%. At the same time, the number of firms spending 3% or more of their budget on marketing rose from 52% to 60%.
Where are firms spending their marketing dollars? The most common marketing spend (63%) is for “networking” — which the report does not define.
Far fewer firms say they are spending money on other forms of marketing, such as social media marketing (38%), search engine optimization (38%), free listings in online directories (31%), blogging and e-bulletins (30%), and video marketing (8%).
The report ends on a cautionary note.
While small law firms have been able this year to “move the needle” and make meaningful progress on their long-standing top goal of spending more time practicing law and less time on administrative tasks, they cannot rest on their laurels, especially in the face of the rapid advancement of AI and constantly changing client expectations.
“It is of critical importance that small firms hold on to their hard-won gains this year in spending more time practicing law and less time dealing with administrative tasks. Avoiding backsliding into previous habits or losing focus because of distractions or shifting attention to other goals will be key.”
LawSites Coverage of Past years of the survey:
2022 Survey
2020 Survey
2019 Survey
2017 Survey
2016 Survey
Thomson Reuters today delivered on its promise to integrate generative AI within its flagship legal research platform, introducing AI Assisted Research in Westlaw Precision, available immediately to all U.S. Precision customers at no extra cost.
It also previewed its coming integration of Casetext CoCounsel as an AI legal assistant across multiple TR products, including Westlaw Precision, Practical Law, Document Intelligence, and HighQ. As of today, Westlaw Precision users can directly access CoCounsel through a button in the navigation bar.
Starting in 2024, TR will begin adding AI-assisted research to Westlaw in markets outside the U.S., starting with the UK, Canada, Australia, and New Zealand.
CoCounsel will also remain available for purchase as a standalone application, called CoCounsel Core.
In addition, TR announced the beta version of its integration of generative AI into Practical Law, a feature that will be released in the first quarter of next year.
All of these developments were previewed yesterday to members of the media during a briefing by TR executives in the company’s Times Square office in New York.
Of all the announcements covered during the briefing, David Wong, TR’s chief product officer, said the launch of AI-assisted research in Precision was “the headline today.”
An answer provided by AI-Assisted Research.
Mike Dahn, SVP and head of Westlaw product management, went even farther, calling the launch “the biggest transformational change in legal research since the move from books to online legal research.”
The key to that change — and what makes AI-Assisted Research different from traditional research — is in how it processes and delivers results. In traditional legal research, the user enters a query and gets a response in the form of a long list of cases, statutes and other resources. It is then up to the user to plod through it all in search of the answer.
With AI-Assisted Research, the user asks a question and the AI delivers an answer in the form of a narrative. No more plodding through cases. The narrative includes footnotes pointing to the cases or other materials on which it bases its answer. Hover over a footnote to see the case and an excerpt. Click on the footnote to see the case, or scroll down on the page to see the full list of results.
Although it can take more than 90 seconds to deliver an answer, it offers users the option of emailing them when the answer is ready.
Because there is a lot of computing going on in the background here, AI-Assisted Research takes longer to deliver its answer than traditional research – anywhere from 90 seconds to three minutes. But Dahn says beta users have so far not put off by this, because they appreciate how much time they save when they do receive the answer.
Behind the scenes, the AI in Westlaw is using retrieval augmented generation (RAG). That means that it uses traditional research methods to pull relevant resources out of Westlaw, which it then sends to the large language model (LLM) to analyze and generate an answer.
Answers contain footnotes to sources relied on. Hover over one to see the context.
For its LLMs, TR is using a combination of Microsoft Azure OpenAI and a direct commercial API to OpenAI’s GPT-4. Importantly, TR says no user data is being directly shared outside the TR environment.
A future release of the AI tool will also allow a researcher to summarize groups of cases. When a headnote lists a number of cases as supporting authorities, the user can click a button to summarize all the cases and make them easier to review.
There are other AI-assisted legal research products already on the market – most notably from LexisNexis, which released its Lexis+ AI last month, and vLex, which also last month released the beta of its generative AI legal research tool. But TR maintains that its is superior to others for two reasons – the quality of its discovery process and the quality of its content.
The quality of the process, TR says, comes from Westlaw’s proven and long-established legal research platform. The quality of the content comes from both the depth of Westlaw’s content as well as its enhancements through KeyCite, Key Numbers, Headnotes and Precision research.
As of today’s launch, the AI research tool draws only from primary law materials in Westlaw, such as cases, statutes and regulations. TR says that it will soon bring secondary materials such as treatises into the mix.
An advantage of this RAG approach is that it minimizes the risk of hallucinations, because it draws its answers from trusted content. But Dahn says hallucinations are not the primary problem with generative AI research. Rather, the bigger problem is accuracy, in that an answer can be free of hallucinations but nevertheless wrong.
“Our focus is on delivering accurate results,” Dahn said. “Hallucination-free is a relatively low bar, but accuracy is more important.”
Even so, the speakers at TR’s briefing uniformly emphasized that AI-Assisted Research is a tool that can supplement and enhance research, but that should not be viewed as a replacement for traditional research. In fact, the welcome page includes this disclaimer:
“AI-Assisted Research uses large language models and can occasionally produce inaccuracies, so it should always be used as part of a research process in in connection with additional research to fully understand the nuance of the issues and further improve accuracy.”
There are also certain types of research that this product does not support. Examples provided by TR include:
Ever since last June, when TR announced that it would pay $650 million to acquire Casetext and its CoCounsel product, the first AI legal assistant developed in partnership with OpenAI, curiosity has been rampant over how TR would integrate the technology.
Now we know more details.
TR said it will be integrating CoCounsel’s legal assistant throughout its legal products, including Westlaw Precision, Practical Law Dynamic Tool Set, Document Intelligence, and HighQ. CoCounsel will also remain available as a standalone application.
CoCounsel offers a set of “skills” that enable users to perform tasks that include prepare for a deposition, draft correspondence, search a database, review documents, summarize a document, extract contract data, and determine contract policy compliance.
Jake Heller, Casetext’s cofounder and now head of product for CoCounsel at TR, said other skills are being developed to address customer-specific needs.
“Today, CoCounsel is the first and best AI legal assistant,” Heller said at yesterday’s briefing. “As an attorney, I can delegate complex substantive work to my AI assistant and expect it to be done at human or superhuman speed and with human or superhuman accuracy.”
Heller demonstrated how the generative AI capabilities of CoCounsel allow it to answer questions that would have been unheard of with traditional search. Upload a set of documents, such as a collection of emails, and ask, “Do these documents contain a joke?” CoCounsel can answer that question, something that search could never have done before.
One new skill launching today in beta within CoCounsel is the ability to create timelines. Users will be able to upload a set of documents and create a chronology of events.
At yesterday’s briefing, Emily Colbert, head of product management, know how and compliance, demonstrated the newly released beta version of a conversational AI interface for Practical Law (PL). The feature will become generally available in the first quarter of 2024.
A user will be able to find PL content by asking a simple query. The interface will read through PL’s resources and provide an answer, together with the sources within PL on which it based the answer.
The conversational interface makes it easier for users to find answers in practice areas or on topics with which they are not highly familiar. It can also sometimes find on-point resources from less-than-obvious topics within PL.
While this beta version uses generative AI to improve customer access to expertise, Colbert said future versions will unlock access to even more content and features within PL, and will eventually transform customers’ workflows by delivering access to PL expertise wherever they are working.
“We’d like to bring that expertise directly into customers’ workflows,” she said.
Cognizant of customers’ concerns about the security of using generative AI, TR says it protects customers’ data through a comprehensive information security management framework and a range of security policies, standards and practices.
Specifically with regard to AI, it says that it expressly prohibits any vendor from retaining or using Westlaw or Practical Law customer data to train its generative AI model.
Its use of generative AI is through dedicated API access via Microsoft Azure, OpenAI and Amazon Web Services, it says, and it uses a risk-based approach to assess vendors, applications and technologies through a variety of testing and assessment methods.
“Thomson Reuters is redefining the way legal work is done by delivering a generative AI-based toolkit to enable attorneys to quickly gather deeper insights and deliver a better work product,” said Wong. “AI-Assisted Research on Westlaw Precision and CoCounsel Core provide the most comprehensive set of generative AI skills that attorneys can use across their research and workflow.”
The fact is that legal research providers are in a horse race to bring generative AI to their platforms. On its face, AI-Assisted Research on Westlaw Precision appears not to be much different than the Lexis+ AI product from LexisNexis or Vincent AI from vLex. They all allow the user to ask a question and get a narrative response with links to the source materials that support the response.
TR argues that the quality and depth of its content and the head start provided by its headnotes and taxonomy of Key Numbers mean it can deliver higher-quality answers. Obviously, the proof of that will be in the pudding, as customers begin to use and evaluation the product.
But what TR has going for it over any of its competitors right now is the one-two punch of AI-Assisted Research and CoCounsel, two complementary AI tools that, together, both facilitate legal research and provide an AI “legal assistant” for other common tasks. As Wong said, that creates “a generative AI-based toolkit” for legal professionals that is useful for legal research and beyond.
Thomson Reuters said today that it will unveil generative AI enhancements to its flagship legal research product Westlaw Precision on Nov. 15 that will build on its $650 million acquisition of Casetext last June.
TR said it is taking a “best of” approach to product development, leveraging Casetext to accelerate the time for bringing its product to market, while also blending content with workflows and expanding beyond legal research to provide benefits to customers across its segments.
With regard to Westlaw Precision, today’s statement said:
“Enhancing its flagship product, Thomson Reuters is set to make a new generative AI skill available in Westlaw Precision later this month. Using the new skill named Westlaw Precision AI-Assisted Research, customers will be able to ask complex questions in conversational language and quickly receive synthesized answers. Leveraging innovation from CoCounsel, the new skill will draw from Thomson Reuters industry-leading, trusted content from across statutes, cases, and regulations to quickly resolve queries that used to take hours.”
The company also said that it would be bringing a chat-type interface to its Practical Law product. Available to customers in January, the new Practical Law Answers will allow users to submit queries in conversational language and quickly receive answers validated by the trusted content of Practical Law.
Today’s statement came as somewhat of a surprise to those of us who cover legal technology, as the company had last week invited us to a press briefing about the Westlaw updates on Nov. 14, with the understanding that the news was under embargo until Nov. 15.
In today’s statement, TR also restated its announcement from earlier this year that it is committed to investing more than $100 million a year in generative AI through a strategy of “build, buy and partnership,” including the Casetext acquisition.
It has been the legal story of the summer. In June, Thomson Reuters announced that it had signed an agreement to acquire the legal research and AI company Casetext for a jaw-dropping $650 million.
The deal, it said at the time, was subject to regulatory approvals and customary closing conditions and was expected to close in the second half of the year.
Now, the deal is done. TR said today that it has closed on its acquisition of Casetext.
“We are thrilled to welcome the Casetext team to the Thomson Reuters family,” Steve Hasker, president and CEO of Thomson Reuters, said in a statement. “Together, we will accelerate breakthroughs in generative AI for the benefit of our customers and markets. We believe this will revolutionize the way professionals work, and the work that they do.”
Jake Heller, Casetext’s founder and CEO said: “Joining Thomson Reuters provides us with an unparalleled opportunity to scale the Casetext vision and accelerate breakthroughs for legal professionals. As we look to unlock generative AI’s full potential to drive greater productivity, more impactful work and increased access to justice, there could not be a better home for our brand, talent and products.”
In October 2020, legal industry veterans Joe Borstein and Paul Stroka set out to change the legal tech sales paradigm by founding LexFusion as a go-to-market representative of a curated collection of companies across major categories of legal technology. As the company nears its third anniversary, Borstein joins me on the LawNext podcast to reflect on its successes and failures and to share where it is today.
In addition, Borstein shares his perspective on the recent acquisition of Casetext by legal tech behemoth Thomson Reuters for $650 million in cash. As it happens, not only was Casetext one of the companies that LexFusion represented, but Borstein is a former executive of Thomson Reuters, where he worked as global director of Thomson Reuters Legal Managed Services (the former Pangea3).
Given that the Casetext deal was driven by its development of CoCounsel, an AI legal assistant powered by GPT-4 and developed in cooperation with GPT’s developer OpenAI, Borstein also offers his views on the impact he sees generative AI having on the legal industry broadly and on the conversations he is having with law firm and corporate legal leaders.
This is Borstein’s fourth appearance on LawNext. His previous episodes were:
On LawNext Podcast: LexFusion’s Joe Borstein and Casey Flaherty on the 2022 Legal Market in Review.
On LawNext: Joe Borstein and Basha Rubin on the Launch of LexFusion.
LawNext Episode 38: EY’s Purchase of Pangea3, with Pangea3’s Joe Borstein and Ed Sohn.
This episode of LawNext is generously made possible by our sponsors. We appreciate their support and hope you will check them out.
Nota, the online business banking platform designed specifically for solo and small law firms.
Paradigm, home to the practice management platforms PracticePanther, Bill4Time, MerusCase and LollyLaw; the e-payments platform Headnote; and the legal accounting software TrustBooks.
Sage Timeslips, trusted by solo and small firms for nearly four decades, offers robust functionality, customizable reports, and the ability to capture time and expenses on the go.
If you enjoy listening to LawNext, please leave us a review wherever you listen to podcasts.
Yesterday afternoon, I had the chance to sit down for about 20 minutes via Zoom with Jake Heller, the founder and CEO of Casetext, to discuss his company’s acquisition by Thomson Reuters in a $650 million cash deal.
What follows is a transcript of our conversation, which I have condensed and edited very slightly for style and continuity.
References to the morning investors’ call are to yesterday morning’s briefing by Thomson Reuters to its institutional investors, which I wrote about here.
A Stanford Law School graduate, where he was president of the Stanford Law Review, Heller founded Casetext in 2013 after stints clerking for 1st U.S. Circuit Court of Appeals Judge Michael Boudin and as an associate at law firm Ropes & Gray.
AMBROGI: Jake, thanks for taking a few minutes. And, again, congratulations. As you know, I’ve followed you guys from the start, and so I’m really excited for all of you on a personal level. But I’m curious about your motivation for entering into this deal. Obviously, you’ve put a lot of blood and sweat into building up this company. On the investor call this morning, the Thomson Reuters executives talked about the strategy behind why they wanted to do the deal. What about from your perspective: Why did you want to do the deal? I’m sure $650 million was a big motivator, but knowing how much you’ve put into the company, I doubt that’s all there is to it.
HELLER: That’s right. We wouldn’t have done anything like this if we didn’t know that it wouldn’t be a good deal not just for the shareholders and the employees, but also for the customers. What really got me excited about this was all the possibilities of what we’re going to build together with our customers.
For one, the number one constraint that we have right now is content. On one hand, we know that providing West’s and Thomson Reuters’ world-class content database behind our legal research capability is going to make that dramatically better, including when CoCounsel does legal research activities for you. Having it do that on Thomson Reuters’ content is going to be an immediate benefit to our customers.
But, working with their content, there is much more that we can do. For example, using Practical Law’s content to make drafting automation tools – tools that are already there in the pipeline for us – just dramatically better. And really, that’s just where it starts. We saw an immediately huge gain for our customers just by working with them on this. And what’s even more exciting to me is it’s clear that TR is going to invest very substantially in this type of AI. They’re taking it as seriously as we are, have the resources to accelerate our development in a way that, working separately, neither of us would be able to accomplish in the same way as we could working together.
For that and many other reasons, as we sat around the table discussing what we hope to potentially accomplish together should this deal close, we all got really excited about that.
One last thing here. This is a company that, through acquisitions, brought on products like Westlaw and Practical Law Company. What we saw with those products is evolving over time to become really amazing, high-quality products that have, in many ways, set the standard for what things should look like in the world of legal tech. So we knew that we’re working with folks with the capacity and ability to deeply invest in something that they care about and make it into something that has broader reach and a bigger impact than we’d be otherwise able to do. We could probably get there eventually, but it would be really hard. And on some things like content, we may never get there.So we just see this as a massive acceleration of what we’re able to do together.
AMBROGI: What about from their side? Several of the investors this morning seemed to be asking about why TR made this acquisition versus developing the technology themselves, or why they acquired Casetext over, say, Harvey or some of the other products out there. Looking at it from TR’s perspective, what is it that you think Casetext brings to TR that they couldn’t have done or wouldn’t have done without you?
HELLER: I don’t want to speculate too much about their own position. But I’ll say, from my perspective, we are doing this better, we’re doing this best. We have just done a really good job of building out capabilities – reliable, professional grade AI in use today by thousands of lawyers to transform their practices. And I believe that our technology, some of the breakthroughs that we’ve had, some of the fundamental capabilities on the platform level and the way that we’ve approached things, is going to massively accelerate their own development of the application of AI to their products. I think they saw that in us.
You’re right. There are a lot of folks out there who are developing against this technology. Harvey is one of them. But as we look at it, there are dozens of new companies coming into the space with AI capabilities. I do think it’s going to be a competitive market – a very competitive market – for having AI help people with legal work. And I think that they just wanted to work with the folks who are doing it at the top of the game. I think that’s at least a piece of it. And I say it with humility because it was really hard to get here, and the team put a lot of work to get to this place. And we know that in order to continue doing it best, we’re going to have to really continue to deeply invest and to not lose any sense of urgency or not get complacent. Now is not the time for that.
AMBROGI: You just mentioned that it’s going to be a really competitive market. One of my concerns is what the impact of this acquisition will be on the competitiveness of the market. Casetext has had a record of being one of the most innovative companies out there and driving innovations that other companies, including TR, have emulated. Assuming the deal closes, once you become part of TR, what happens? Does that make the market less competitive? Do you have any fear that you will be less innovative or less able to roll out the kinds of innovation you’re known for?
HELLER: I wouldn’t be doing this if I thought that might be the case. You know me. You’ve been covering us for a long time. If I was worried about that, we wouldn’t be walking into this deal. Frankly, we had too many other opportunities to do too many other things that, if we took that out of what we’re doing, I’d be really concerned. To the contrary, it’s quite the opposite. They’re bringing over the entire Casetext team, and they’re investing very heavily in this. We expect the amount of resources that we’ll be able to expend on innovation – not just monetary, but know-how and customer relations and content and everything else that TR provides – is going to accelerate rather than decelerate our rate of innovation. That’s exactly the intention.
I respect that, from the outside, some of these deals kind of look the same, right? But, being on the inside of it, we have gotten very confident, extremely confident, around what this will mean for our pace of innovation and what we’ll be able to accomplish, and we are really encouraged by the way that their leadership team is thinking about this and is thinking about what the next stage looks like.
I don’t think that us being a part of TR, accelerating our pace of innovation, is going to take out the competitive infants of the market. In some ways, I think, it’s going to raise the bar for the whole market about what can be possible, what can be done by this kind of technology. As you saw, just yesterday, Relativity said it is getting into generative AI. Harvey is obviously raising a lot of capital. EvenUp just raised $50 million in capital to provide something for plaintiffs’ attorneys. I think you’re going to see a ton of activity because this is one of the most exciting, most important moments in the history of legal technology, if not the most exciting moment of legal technology, I don’t think you’re going to see a bunch of people waiting on the sidelines or unable or unwilling to get in the fray. You’re seeing the exact opposite.
AMBROGI: Do you know yet what the roles will be for you and cofounders Pablo Arredondo and Laura Safdie?
HELLER: Obviously, some of the details have to be worked out as we reach closing. We are all joining – us and our chief technology officer and our chief revenue officer – as with the rest of the company. All of the employers are also joining. I do know that I’m going to be really focused on product and really focused on innovation, as will Pablo and as will my chief technology officer. I know that Laura is going to be really involved in managing the integration and running the operations of the business. And we know that they’re keeping a lot of the group together as part of one organization. They don’t want to change what’s working. Don’t fix what ain’t broke. But what they do want to do is put more resources in it. So I’d expect to see more people working on the same core technology with more resources, more content, more everything behind it after the transaction closes. That’s part of why we’re really excited about it.
AMBROGI: Part of the discussion at the investor call this morning was around the price tag, and a lot of the talk in the industry has been around the price tag. The Thompson Reuters executives on the call this morning made some allusions to the fact that your revenues are relatively modest, particularly in comparison to TR’s. And your product, your AI product, CoCounsel, is still a relatively new product. It was just brought to market, just released formally, four months ago. So how do you explain this price tag? How does this price tag make sense for the legal tech market?
HELLER: It’s a great question. I think part of it we brought our product to market incredibly quickly, and I think that’s something that they saw. I think it’s a competitive market to fund and acquire businesses. We know what we were getting offered from the likes of venture capitalists and so on, and I think that drove an understanding of what the market believes in terms of our potential. But I think the bigger picture here is the price tag reflects the upside of what they can do together, working with us. I think they saw that if you combine our technology and their content, their know-how, their customer reach, that the opportunity here was enormous and presumably even bigger in their eyes than the amount they’re investing in acquiring the business and will be investing in growing the business. So I think it speaks to their optimism for what we can accomplish together as a company.
AMBROGI: What happens with CoCounsel? To what extent have you had conversations with them about that? Do you see CoCounsel continuing as a standalone product or do you see that it becomes merged into however Thompson Reuters ends up incorporating this technology?
HELLER: What I think you’ve seen with Thompson Reuters and their track record is they’ve been very good about supporting products, oftentimes independently from the main product. A great example is Practical Law, that one can subscribe to hypothetically without being a Westlaw customer, without being otherwise a TR customer. We know that folks who are subscribing to products like CoCounsel are going to continue to have access to it exactly as they have, continue to subscribe to it exactly as they have. And my hope and expectation is a lot of the core underlying technologies we’ve been developing will, as their content and as their information improves CoCounsel – for example, CoCounsel leveraging Practical Law and Westlaw data to get smarter and better at what it does – I hope that some of our technology also bleeds over and makes the core Westlaw experience better and the core Practical Law experience better.
Something we’ve talked about extensively is that a lot of what we’ve done with our platform – some fundamental breakthroughs in the way that we handle this AI – also applies to tax and accounting and compliance and regulatory, and news through Reuters. So I expect that while they’re helping make our product better, we are going to help have an influence and an impact across all the professions they serve and the news. And that’s also something that got our team really excited – the opportunity to have such a broader impact based on the work they were doing and extend that influence across so much else.
AMBROGI: I know your time is short. Anything else that you’d like to say before we wrap up?
HELLER: You know, I’ve been doing this for a while and, as you know, there are certain things that we really care about as a company, right? We care about innovation. We care about our customers deeply. We really care, especially right now, about getting this AI technology right, about doing the best we possibly can for our customers during this time of what we believe is a shift in the history of legal technology. For us – and you know us – for us, the most important thing about this deal is whether we can continue that pace of innovation, about knowing that we can serve our customers better with this combination, that was a major driving force behind this.
If you were to take the mood within Casetext of the employees, part of why people are so excited about this – and a lot of them just found out for the first time themselves yesterday, because we had to hold that information close to the chest – is that we all know that we are in a better place together and we’re just so excited about that. I hope folks give us a shot to see what that looks like because I think they’re really excited and really impressed.
Our mission is around empowering attorneys to make their practices so effective and so efficient that they can, in turn, provide their clients with justice, right. Something we think a lot about on that point is whether we are serving our mission well for not reaching as many people as possible? There are very few companies in the world that can help accelerate the adoption of this technology, and the responsible adoption of this technology with the right training and guidance and set up, like TR. It’s a unique opportunity to help accelerate our mission.
AMBROGI: Again, a big congratulations to you, and thanks for taking the time out of what I’m sure is an extremely busy day to talk to me.
HELLER: Anytime.
Compared to Thomson Reuters, Casetext is a relatively small company with relatively minimal revenue, and one that effectively refocused its product strategy over the past year, launching its generative AI product CoCounsel just four months ago.
So why did TR shell out a record $650 million in cash to acquire Casetext? That was the question on the minds of several of TR’s institutional investors during a call this morning at which the company briefed them on the deal, announced late last night.
The answer, effectively, boils down to: It’s all about the AI.
The briefing was conducted by TR CEO Steve Hasker, Chief Financial Officer Michael Eastwood, and Chief Product Officer David Wong.
During the call, several investors explicitly asked the three executives why TR had decided to acquire Casetext rather than build its own generative AI.
Their answer was that Casetext’s early access to OpenAI’s GPT-4 model and its experience in working with legal data gave them a head start in developing the know-how and experience to engineer generative AI products, and that experience will help accelerate TR’s own generative AI strategy.
Casetext was one of only a few companies to which OpenAI provided access to the advanced GPT-4 model as early as the first half of 2022, they said.
“They’ve had a chance to be able to work with OpenAI to understand the model, tok be able to design use cases, and to become familiar with how best to prompt the model for effective application within their domain,” said CPO David Wong.
Wong also cited Casetext’s development of a set of eight prompts designed to enable GPT-4 to perform taks specific to the legal domain, such as review documents, summarize documents, and search a database. He said that several of Casetext’s prompts are unique and are ones TR had not even started to develop.
Investors also asked why Casetext as opposed to other generative AI companies in legal such as Harvey, which the law firm Allen & Overy deployed firmwide last February.
Again, the TR executives cited Casetext’s head start in working with GPT-4, its established access to and expertise in legal research data, and the uniqueness of its prompts.
“What’s unique about Casetext and how they have built CoCounsel is they’re one of a very few set of companies out there that also have access to data,” Wong said. “They had access to, again, the historical Casetext research data set, and so they had the benefit of both the early access to GPT-4 as well as experience in how to combine it with an authoritative data set, albeit again, a smaller, less sophisticated one than we have.
“I think that made them unique versus others in the market that are just using LLMs kind of by themselves. That meant that their team, their know how, their experience, could be much more easily translated into our own teams, where we’re, again, one of the small number of companies that have a really authoritative data set and the talent and know how to apply generative AI.”
Several investors asked about Casetext’s revenues and pricing, but the TR executives put off explicit answers about any of that for now, promising more details when they issue their earnings report later this year.
“While Casetext brings limited initial revenue, we expect it to grow rapidly and be accretive to revenue growth over the next few years,” Hasker said.
Outlining their broader strategy around the deal, the TR executives said that they believe the CoCounsel product will accelerate their ability to bring generative AI’s capabilities to their product portfolios and markets.
“Marrying CoCounsel’s technology with TR’s authoritative content and insights to bring significant productivity benefits for the legal industry,” they said.
The all-cash purchase price of $650 million was funded by proceeds from TR’s recent sales of its stake in the London Stock Exchange. They expect the transaction to close in the second half of this year.
They outlined four key strategic benefits to the deal:
They described the acquisition as part of a three-pronged “build-partner-buy” generative AI strategy to build solutions of their own, partner with trusted companies such as Microsoft, and buy leading AI powered solutions.
“We believe CoCounsel is the most advanced legal AI assistant on the market today and see the combination of its leading technology with Thompson Reuters’ authoritative and trusted content and insights providing significant value for the legal industry,” said Wong. “The combination of Casetext with Thompson Reuters offers a compelling strategic fit that will provide meaningful benefits for our customers and strong long-term value creation for our shareholders.”
Rumors had been swirling that Thomson Reuters was about to acquire Casetext, and they proved to be true.
Just before 9 p.m. Eastern time today, TR said it had signed a definitive agreement to acquire Casetext for $650 million cash.
“The proposed transaction will complement Thomson Reuters existing AI roadmap and builds on its recent initiatives, including a commitment to invest more than $100 million annually on AI capabilities, the development of new generative AI experiences across its product suite, as well as a new plugin with Microsoft and Microsoft 365 Copilot for legal professionals,” TR’s announcement said.
The acquisition unites a long-established giant in legal research with a company that has proven to be a feisty upstart since its founding in 2013. As I wrote of Casetext just two weeks ago, as it marked its 10th anniversary, “From the very start, Casetext was an innovator, always pushing the envelope for what a legal tech product and company should look like. By doing that, it has been always a trailblazer, never a follower.”
Earlier this year, it launched CoCounsel, an AI legal assistant developed in partnership with OpenAI and using the latest version of OpenAI’s GPT large language model.
Even before launching CoCounsel, Casetext had already launched the powerful neural net search technology AllSearch and had pioneered products such as Compose, to help lawyers draft litigation briefs, and, in 2016, CARA, the first product to use AI to analyze briefs, which spawned a generation of copycat products, including by Thomson Reuters.
“The acquisition of Casetext is another step in our ‘build, partner and buy’ strategy to bring generative AI solutions to our customers,” said Steve Hasker, president and CEO of Thomson Reuters. “We believe that Casetext will accelerate and expand our market potential for these offerings – revolutionizing the way professionals work, and the work they do.”
“For the last ten years, we have harnessed the power of AI to build products that elevate the practice of law and enable attorneys to serve more people’s legal needs, with the ultimate goal of increasing access to justice,” said Jake Heller, CEO of Casetext. “Joining Thomson Reuters is an incredible opportunity to advance our mission and the field of generative AI solutions exponentially, not only for lawyers but across professions, ensuring this revolutionary technology can benefit as many people as possible.”
TR said that closing of the transaction is subject to specified regulatory approvals and customary closing conditions and is anticipated to occur in the second half of 2023.
I am of mixed emotions about this. While I am thrilled for Casetext’s founders, who have worked hard for 10 years to create one of the most innovative tech companies in legal, I fear for the implications of the loss of a feisty competitor. Not only has Casetext been innovative, but it has launched products so original that other companies have had to scurry to catch up with them.
With Casetext now swallowed up by the industry giant, will that spirit of innovation and originality be lost? I sure hope not.
As other legal technology companies have rushed to release products using generative AI, one of the largest, Thomson Reuters, has been unusually silent. It broke that silence slightly two weeks ago, when it announced plans to invest some $100 million a year in AI, including incorporating generative AI across its flagship products in the second half of this year.
Today, TR provided more explicit details regarding its plans for incorporating generative AI and chat into its legal products during presentation for the media by David Wong, chief product officer.
Specifically, it revealed that it will be incorporating chat functionality into its legal research and workflow products by the second half of the year, and that it will begin inviting select customers to test prototypes even sooner.
TR also announced a partnership with Microsoft by which its AI-powered products will be integrated within Microsoft 365 Copilot, Microsoft’s tool for providing advanced AI capabilities across its productivity suite.
TR and Microsoft have collaborated to develop a contract drafting solution that will combine TR’s software and content with Copilot for Word. A lawyer or legal professional working in Word will be able to draw on TR resources such as Practical Law and Westlaw as they draft.
Wong said this will be the first legal product to integrate with Copilot and one of the first of any third-party integrations.
With regard to TR’s primary legal products, its planned integrations are all still early prototypes. But the company offered previews today to members of the media. Among the integrations TR previewed:
Chat within Westlaw Precision. The integration of chat within Precision will appear as a two-panel screen, with a familiar chat interface on the left, where a user can ask questions and view the AI’s responses, while on the right the user will see the references that support the answers.
Chat within Practical Law. With a feature tentatively named Practical Law Dynamic, a user will be able to engage with Practical Law content through a chat interface. Similar to chat in Precision, the prototype shows a dual-pane interface with the chat on the left and references or related data on the right.
Generative AI within HighQ. Wong said that TR sees opportunities to use generative AI within HighQ to enhance automation of workflows and pulling of information from different sources.
Wong said that TR will be providing more details and more updates in the coming weeks and months.
As product prototypes become available, he said, selected customers will be invited to test them and help TR learn from their experiences.
He said that chat functionality will be integrated into TR’s legal research and workflow products in the second half of this year.
The Copilot scheduling is contingent on Microsoft’s timing, he said, but TR will be at the forefront of those releases.
On a broader scale, Wong said that TR is committed to leading the development of generative AI for the legal market, and is well positioned to do so given its history of developing AI in legal products.
Login