Top A.I. Startups in the Future of Legal Technology
The Futuristic Lawyer gives his analysis: A guest post by Tobias Jensen
“The world's best AI legal assistant” according to Stable Diffusion
Welcome Back!
Today I have a guest post from a Legal advisor and copyright and A.I. ethics researcher I really admire. Tobias hails out of Denmark, and we have had many lively debates about A.I. topics in recent months.
Some analysts believe Generative A.I. could radically alter the practice of law. Whether you believe that or not, there are many interesting A.I. startups at the intersection of law and Tobias narrowed it down to what he thought were among the most legit ones. OpenAI has invested in Harvey as of November, 2022.
📶 From our sponsor: 📶
🗣 Don't settle for mediocre meetings | Poised: the AI-powered communication coach
Get real-time feedback to enhance your clarity and confidence, reduce unproductive meeting time, and make a more significant impact with every interaction. TRY POISED to become a clear and confident communicator!
Use code: AISUPREMACY30 for 30% off.
From the Author
The Legal Copyright Battle Against AI – Introduction to EU’s Requirements
Tobias is author of The Gap, and a legal expert in A.I.
Upgrade your membership in this community, I’m going to be increasing my prices on September 1st, so get it while it’s low.
By
In a Goldman Sachs report from earlier this year, it was estimated that 44% of legal tasks could be automated in the future by generative AI. I strongly believe that the future is already here now. A new generation of AI-based legaltech solutions is out on the market, and they will reshape how lawyers work and how we think about law.
Many, if not all, of the world’s leading law firms are already using generative AI tools as a part of their daily work. Over the coming years this development, which was supercharged by the introduction of ChatGPT and GPT-4, will only gain more momentum.
In this post, we will take a closer look at three of the leading startups in the future of legaltech: Harvey, Casetext, and Luminance.
1. Harvey
Location: San Francisco, US
Valuation: $150M
Harvey was founded in 2022 by Winston Weinberg, a former antitrust and securities lawyer, and Gabriel Pereyra, a former AI researcher with experience from Meta, DeepMind, and Google Brain. Earlier in 2023, Harvey raised $21M in a Series A funding round led by Sequoia Capital. The funding round included participation of OpenAI through the company’s startup fund.
Harvey presents itself as “Generative AI for Elite Law Firms”. They are cooperating with large international companies like UK-based law firm Allen & Overy and PwC. According to Sequoia Capital, more than 15.000 law firms are on the waiting list to try Harvey’s AI technology.
The Harvey AI is a kind of ChatGPT for lawyers. It is built on top of OpenAI’s GPT-4 and fine-tuned specifically for legal tasks. Co-founder Gabriel Pereyra explains in an interview with TechCrunch: “Instead of manually editing legal documents or performing legal research, Harvey enables lawyers to describe the task they wish to accomplish in simple instructions and receive the generated result.”
It sounds like Harvey is aiming to replace lawyers with chatbots. But Pereyra insists that that is not the case: “Harvey will make lawyers more efficient, allowing them to produce higher quality work and spend more time on the high value parts of their job.”
In Sequoia Capital’s announcement, Harvey’s mission sounds even more ambitious: “Harvey is the first and best instantiation of a new breed of company: the AI super app. Harvey is in the business of giving people superpowers. First, lawyers. Next, professional services. Eventually, all of knowledge work.”
Hmm. There is not any public information available about how the underlying technology works or what Harvey’s user interface looks like. Neither is it possible to take a trial or find an online demo. Apparently, the Harvey team has decided to go with OpenAI’s closed approach to AI deployment for the sake of gaining a competitive advantage.
As a legal professional and a human being, I find Harvey’s closed approach disturbing. And bear with me for a second, while I put on my tinfoil hat. No company should be allowed to develop and apply powerful AI behind closed doors. Along the same lines as companies are not allowed to clone humans or develop nuclear weapons. Harvey’s lack of transparency should be illegal, and when EU’s upcoming AI Act enters into force, it rightfully will be, at least in Europe.
Editor’s Note:
If this topic is of interest to someone you know, why not share it:
You can also share the Newsletter and unlock free paid complementary subscriptions by doing so.
Tobias Jensen (Futuristic Lawyer) and I share many of the same interests in topics. In the past few years he’s specialized more on blockchain, distributed ledgers and Web3 at the intersection of law, but lately has also become very interested in A.I. topics. The EU AI Act might change everything.
Subscribe to learn more:
Let’s continue our deep dive.
2. Casetext
Location: San Francisco, US
Valuation: In the range of $50M to $100M
The San Francisco-based Casetext was founded in 2013 by three litigation lawyers, Jack Heller, Laura Safdie, and Pablo Arredondo. The founders each have experience from different areas of the law. Of all the vendors in the Legal-AI space, Casetext is one of the most established players and arguably provides the most advanced product. By all measures, Casetext is a big deal.
Casetext has raised a total of $64.3M over 7 funding rounds from top-tier VCs such as Union Square Ventures, Canvas Ventures, Y Combinator, and Touchdown Ventures. In their latest Series C funding round announced in January 2022, the company raised $25M. Recently, sources have revealed to Business Insider that Casetext is in talks to be acquired after a hot funding round, speculatively by Thomson Reuters, which owns the legal research platform Westlaw. A potential deal could value Casetext in the hundreds of millions. However, a deal is not guaranteed and could end up not happening.
Casetext offers a product called CoCounsel which is an AI legal assistant, or essentially an advanced, AI-powered research tool. It can help with document reviews, summarization of text, extracting certain data from documents, searching through databases, and answering questions. You can watch a demo here.
As with Harvey AI, CoCounsel is built on GPT-4 and customized for the legal industry. The two AI products provide competing services, but I am much more comfortable with CoCounsel. Casetext shares all relevant details about its products on the website, offers a 7-day free trial, and importantly a satisfying security overview.
Casetext services more than 10.000 law firms including DLA Piper and other large, international firms. In terms of pricing, full access to CoCounsel costs $400/month, while a “Basic Research” subscription with access to Casetext’s search tools and comprehensive database of US state and federal law costs $110/month.
Editor’s note: Today Thomas Reuters announced that it’s acquiring Casetext, for $650 million.
From the Author
The Legal Copyright Battle Against AI – Introduction to EU’s Requirements
Tobias is author of The Gap, and a legal expert in A.I. and researcher in A.I. ethics and emerging technologies.
3. Luminance
Location: Cambridge, UK
Valuation: $100M
Luminance was founded in September 2015 by a group of mathematicians from the University of Cambridge. The company has raised a total of $23M in funding over 5 rounds.
Luminance works as an advanced, AI-powered contract management platform. Among its large suite of functions, Luminance can be used to generate contracts from scratch that comply with internal company standards and provides guidance along the way that even non-lawyers can follow. In contract negotiations, Luminance can automatically color-code the received document in Microsoft Word after which parts are acceptable and which parts require further review. Luminance can also provide insights within its contract repository, help to visualize key data points, and flag areas of business risk and opportunity.
Last month, Luminance announced the launch of its specialist legal large language model ‘Ask Lumi’. Opposed to Harvey AI and CoCounsel, Ask Lumi is not designed to answer general questions about law, but specific questions and queries regarding drafting and negotiation of contracts.
According to the one-liner from Luminance’s homepage: “Luminance is the world’s most advanced AI for the processing of legal documents, streamlining operations and delivering value business-wide.” Luminance services +500 companies in +60 countries in +80 languages and has so far been used to analyze +150M documents. Allegedly, Luminance has all of the Big Four consultancy firms and over one-quarter of The Global Top 100 law firms as customers.
Although Luminance strikes me as a remarkably powerful tool, I remain skeptical about whether Luminance is indeed the world’s most advanced AI for processing legal documents. This claim may have been valid three years ago, but with the recent quantum leaps in generative AI with GPT-4, utilized by Harvey and Casetext, I have my doubts about Luminance’s leading position. However, I could be wrong! If I had to choose one AI tool for drafting, negotiating, and managing contracts, it would be Luminance.
Recap
Harvey has received quite a lot of attention recently as one of the first tools built on GPT-4 with backing from OpenAIs startup fund. The company is on an ambitious mission to give lawyers, followed by all other knowledge workers, superpowers with AI. The industry-leading companies A&O and PwC are already using and helping to develop Harvey AI. So far, Harvey is embedded in secrecy and a mysterious aura, since the public doesn’t know how Harvey works or what it looks like. Considering the impact Harvey could potentially have on the job market and eventually the rule of law, I find this closed approach concerning.
Casetext has been around since 2013 and is arguably the most advanced legaltech company on the market today used by more than 10.000 law firms. Casetext's flagship product CoCounsel is a legal research tool built on GPT-4’s advanced generative AI capabilities, destined to cut down millions of billable hours.
Luminance is founded in the UK, far away from Silicon Valley, and developed by mathematicians, not by former litigation lawyers or CTOs, like its competitors. While Casetext is at its core a legal research tool, Luminance is at its core an advanced platform for contract management. Luminance can be used by lawyers anywhere in the world, whereas other leading AI legal tools are based in the US market and on the common law system. I am convinced that Luminance will play a significant role in reshaping the legal work of tomorrow.
Thanks for reading!
Read More from This Author:
If you want to be a guest post contributor in this Newsletter, read this.
Addendum: Reuters Acquires Casetext
The day this post was going live, Reuters said it is acquiring Casetext for $650 million, a Y Combinator-backed legal tech startup. The ten year old startup initially focused on creating both a community for attorneys to share knowledge and a service to give users free access to legal texts annotated by lawyers. But the company, per TechCrunch, later pivoted, embracing AI and ML to build automated workflows and tools for legal teams.
Casetext’s flagship product is CoCounsel, which taps AI to review documents, help with legal research memos, prepare depositions and analyze contracts.
Reuters says it’s set to invest more than $100 million annually on AI capabilities as a new plugin with Microsoft and Microsoft 365 Copilot for legal professionals. There was another acquisition of note recently in the Generative A.I. space by Databricks that I found rather notable.
Until next time. Read Tobias on the Gap, and let us know what your thoughts of this topic and what implications you see of Generative A.I.’s impact on law and the legal sector?
 | A guest post by
|
Very cool summary! Also, Tobias- your work is at the juncture of a lot of things I enjoy reading and writing about (especially the very sticky AI copyright stuff we're facing today and tomorrow). You have a new reader.
How did you find this sponsor, is there a platform of some sort to get into contact with potential advertisers?