I’m convinced A.I. is going to eat the software world in ways big and small.
Everyone has heard of ChatGPT at this point. It's the fastest-growing software product of all time - faster than every social network or search engine that we’ve ever seen, or anything, really.
It hit a million users in only five days, according to Statista. By comparison, Netflix took 3.5 years, and Instagram took 2.5 months, despite arguably having far more mass appeal than a chatbot that generates text.
Why? Quite frankly, it’s cool: you ask the robot to tell you something, and it gives you a wholly unique answer in seconds. It’s a chatbot version of talking to a Star Trek computer, though not quite as accurate, to put it mildly.
In the few years that the GPT systems have been available for public usage, they have come a very long way: from producing very hard to read, spammy, inaccurate content to being able to produce a 1,000-word explainer on the fine points of dividing 401k plans in a divorce accurately.
At the current rate of development, it’s not a stretch to imagine AI systems becoming the dominant form of research within a few years, not to mention a critical value-add for other software, such as legal drafting software or even your email platform. Heck, OpenAI's GPT4 passed the bar exam.
Both Microsoft and Google are in the midst of launching their office suite platforms, giving everyone an A.I.-assist on their email, document drafting, and slide decks.
We’ve talked a bit about AI tools for lawyers, but as many of these tools reached maturation, we thought it would be a good time for a full-on map of lawyer-dedicated AI tools: from drafting aids to eDiscovery robots. And to start, we’ll look at yours truly, Spellbook, the first generative AI drafting assistant for lawyers that lives in Word. After that, we’ll continue with tools for litigation analytics, drafting, and other vital parts of the practice of law. Consider this your roadmap to the current state of A.I. in law, with the understanding that this changes by the minute and we’ll try to keep the post updated accordingly.
If your business involves drafting or reviewing contracts, Spellbook is going to be your best friend. Why? First of all, it lives in the sidebar of Word – where all lawyers live – ready to accelerate your drafting.
Second, while many experienced contract lawyers already know to look for obscure clauses, such as a choice of law clause, it can’t hurt to have an A.I. assistant looking over your shoulder, finding clauses that you may have missed. And, when it does find something you missed, it can even propose language for the clause.
I’ve also used it on my core practice area, qualified domestic relations orders, when reviewing templates suggested by others, and it has caught missing clauses there as well, saving me a ton of time comparing my standard templates to opposing parties’ proffered drafts.
With over 1,700 legal teams, in 50+ countries, it’s becoming the go-to AI for lawyers as one of the fastest growing solutions in the space.
While not widely available yet, Harvey is riding a wave of hype. How so? How’s an $80 million investment that valued the startup at $715 million?
Harvey promises to allow lawyers to use natural language to analyze and generate documents. One example, provided to TechCrunch, is, “Tell me if this clause in a lease is in violation of California law, and if so, rewrite it so it is no longer in violation.” While they initially plan on being a product for lawyers, as this tech matures, one can easily see this being a consumer solution that’s far more useful than the legal information websites or will generation CD-ROMs of yore.
What is the product like to use? Well, despite the hype and funding, very few know. According to sources, they have a few large law firm customers. Given their target market and limited availability so far, this may be an aspirational product for most of us lawyers, sort of like a Gulfstream jet, rather than a tool you can use today.
Lex Machina was into A.I. before it was a buzzword – they won an award back in 2019 for their use of machine learning and other integrated technologies to build a litigation analytics platform. Essentially, their systems use AI to categorize and analyze court documents to provide trends on judges, courts, lawyers, and parties in areas like case outcomes, findings, damages, and more. Lex Machina is part of LexisNexis, so you may have access to it if you’re a user of the Big Red Machine for your legal research.
For most people, Lex Machina analytics will not have a daily use case in their law firm. However, for larger litigation firms or venue shoppers, one could see this being very valuable for knowing the opposition, the courts, and the trends that point towards being plaintiff-friendly or not.
Luminance is an end-to-end AI platform covering everything from eDiscovery document review to reviewing contracts, to drafting new contracts. It even colour-codes clauses by acceptability, which is a pretty nifty idea. It seems to be mostly focused on mass document review, rather than drafting contracts clause by clause, though it does purport to do that.
The biggest issue with Luminance is that it is an enterprise-grade tool with mystery pricing - their website doesn’t even address the issue. Because of this, one imagines it is likely out of the reach of most small firms and solo practitioners.
Kira is another AI contract review platform that uses “patented machine learning” to analyze contracts, categorize clauses, and identify risk areas in contract language.
Much like Luminance, it seems to be more focused on reviewing documents at scale – imagine having the A.I. review thousands of contracts and summarize the presence of different types of clauses across those contracts - rather than being a one-on-one drafting assistant, like Spellbook.
Both Litera and Kira have opaque pricing that makes it difficult to ascertain whether they are a fit for a small firm or are better suited to large litigation teams.
Evisort is a contract lifecycle management tool, with integrated AI, that helps manage contract drafting, review, approvals, and more.
Suppose you’ve ever worked in-house at a company and had fifty people across marketing and sales all trying to juggle purchase orders and SLAs via email, complaining because legal isn’t reviewing them fast enough (while renegotiating terms in your standardized agreements with outside parties). In that case, you know how important a CLM tool can be for monitoring these contracts and ensuring they get the appropriate levels of review and approval before heading out into the market.
Need a contract lifecycle management tool? Yes, there are a lot of these, and they are all in an arms race to integrate AI into contract review and drafting.
Ironclad made legaltech headlines by being one of the first to integrate GPT3 into their platform, allowing users to review contracts according to a company’s pre-approved toolbox of contract terms and clauses.
One would be well-advised, if they are in the market for a contract lifecycle management tool, to demo each of these and compare the features and costs before making the leap.
Redline. Redraft. Resubmit. Repeat.
Much of the average in house lawyer’s time is spent reviewing the same old contracts over and over. And, too much of the average marketing, biz development, or salesperson’s time is spent nagging legally to get their papers back to them.
Much like Ironclad, Lawgeex addresses this problem with A.I.-based contract review and redlining. It also lets you define your own contract playbook, so those redlines and reviews will be determined by your company’s standard terms, not by some arbitrary A.I. mind’s whims.
Everlaw is a long-time leader in eDiscovery software, and much of its industry-leading feature set is powered by machine learning, which is a form of A.I.
Essentially, the software gets smarter as you work with it more, categorizing and clustering the heck out of piles of eDiscovery PDFs.
Contrast this learning and categorizing AI that is more about making existing content more manageable with generative AI, like Spellbook, that is more focused on generating new clauses and content.
DocketAlarm is now a legaltech veteran (and is owned by Fastcase), but it isn’t resting on its laurels. The docket tracking platform now includes GPT 3.5-based summaries of litigation briefs and activity, per Law.com, which is a nice automated way to see what the filings you’re perusing are about without buying every filing and reading it yourself.
Clearbrief is a legal tool that uses artificial intelligence to help lawyers write better briefs and showcase the facts and law behind their arguments. It works as an add-in to Microsoft Word and allows users to find relevant evidence from discovery or any PDFs they upload, check the accuracy of their citations, generate a table of authorities, hyperlink their sources, and share an interactive brief with the court and opposing counsel. Clearbrief aims to save time, improve quality, and increase persuasiveness of legal writing.
BriefCatch is a drafting and editing tool for lawyers. It is designed by a legal writer for legal professionals and works on Windows through Microsoft Word (or through Parallels on Mac). In addition to enforcing plan language writing standards, it will compare your writing to notable legal luminaries, like Justice Kagan or Paul Clement.
All three of these research platforms purport to use AI to improve their legal research platforms.
In particular, Casetest started using AI-powered search features a few years ago, which means that it not only crawls citations or keywords like we always have, but also uses AI to comb through the details, court history, and jurisdictional authorities of your actual matter to find more and better results, quicker.
Instead of looking for keywords or bluebook citations, it is looking for concepts and semantic meanings. Compared to a study conducted in 2018, which showed 21% more relevant results, 24.5% faster research times, and much happier lawyers, one can only speculate on how much better the AI model has become since then.
The “Wexis” duopoly also claims to have incorporate AI into their platforms to improve performance.
EvenUp is a startup that uses technology and AI to assist lawyers in personal injury cases.
It uses AI to help lawyers and victims evaluate injury claims, review medical documents, produce medical summaries, and calculate damage estimates.
After a $10 million seed funding round in 2022, the startup locked up a Series B round at a $350 million valuation in early April. EvenUp claims to have over 200 law firms as customers, and to have generated over $350 million in claimed damages.
AI is an unfathomably fast moving field. In just the last year, it has gone from niche toy to invaluable tool that is being incorporated into everything from your browser to search engine to office software.
Even in the past few weeks, developers have pushed AI into thinking on its own to chain together requests, so that it “thinks” on its own about solving problems, rather than just spitting out text or image answers to a single prompt.
With all that being said, the above list is meant as a roadmap to many of the legal-focused AI tools out there. We will keep it updated as we discover more tools and as AI continues to exponentially grow in its abilities. And who knows: maybe someday soon, the AI will curate this roadmap and update it for us.
In the meantime, if you want access to an AI tool for lawyers that is available now, set up a demo with the Spellbook team and accelerate your drafting today.