The public release of ChatGPT by OpenAI in late 2022 brought with it a profound mix of emotions and excitement not seen since the advent of the internet, and some even more dramatic would suggest, since Gutenberg’s printing press. Artificial intelligence (AI) technologies, including ChatGPT’s “large language models,” or LLMs, are potential catalysts to change both society generally, and law specifically, because in both cases they vastly increase the quality of access to various areas of knowledge as well as the speed of using that access. With regard to law, despite the best efforts of the occasional Luddite attorneys, we expect that AI will quickly—over the next five to 10 years—improve the practice of law for client and practitioner alike. This article provides a high-level overview of some currently available tools and use cases, particularly in M&A transactions, and discusses how they are implemented and raises issues that do (and may soon) arise.
Due Diligence
Due diligence is vital for the buyer to learn about a prospective seller’s business, but it is expensive, time-consuming and, some would argue, boring. AI can help with at least the expense and time costs as of today. Software programs including Kira, eBrevia and Ansarada can perform this high-volume repetitive task via AI. First, the lawyer simply uploads diligence files to the program and inputs the types of provisions he or she would like summarized. From there, the program makes it a snap—within minutes, even seconds, the program “reads” the files and produces reports identifying applicable provisions such as change of control, term and termination, all in chart form. The makers of these programs advertise the high accuracy rates of the software’s review compared to traditional methods (i.e., an associate reviewing a contract at 2 a.m.). These programs possess an amazing ability to review and synthesize information and, as impressively, are capable of “learning” new provisions from time to time or being taught which provisions do or do not apply. Human lawyers still have some role in this analysis in guiding and checking results and then summarizing them in the form of diligence memoranda or schedules, as the client may deem appropriate, but ultimately the AI programs can do an initial, increasingly accurate review of a high volume of contracts in far less time.
Closing Documents
Another facet of transactions that can monopolize legal professionals’ time is coordinating the closing process, including gathering signatures and producing closing binders. Gathering signatures has in recent years become easier as programs like DocuSign (not really AI, but software nonetheless) has made soliciting signatures much easier. Regarding closing binders, AI programs on the market now, like DealCloser, bundledocs and Closing Folder, can likewise prepare binders in a matter of minutes. Here, the associate will manage the various agreements as the deal progresses toward closing, so that once all documents are finalized and executed (indeed usually through DocuSign or its progeny), it requires little more than pushing a button to produce an interactive table of contents linked to all applicable documents (and again, of course, confirming that the AI generated the correct closing binder). Here again, AI creates substantial fee and time savings.
Purchase Agreements
Lest we think AI is only coming for the junior associates and paralegals, it is also emerging as a drafting tool for more sophisticated documents, like purchase agreements. Software such as Spellbook uses GPT-4 (the technical descriptor for ChatGPT) integrated into Microsoft Word to make drafting suggestions by comparing contracts with precedents, looking for conflicting terms, fixing formatting issues, identifying potential risks and omissions, and creating redlines. Again, this still requires humans who understand business priorities, client preferences, leverage, and other elements that are more art that science to synthesize and analyze the software’s suggestions. AI drafting tools are newer than diligence or closing binder tools—think the last year rather than the last five—and as such, are not yet ubiquitous among law firms. However, these tools too represent the real near-term likelihood that AI drafting can increase the quality of work product using this technology and improve the speed of drafting.
Research
The classical—and true—perception of lawyers is that they are, as a class, trained to synthesize complex information well and apply law to facts to help clients make decisions. Again, AI tools are developing to improve the speed at which this fundamental task is done. The emergence of ChatGPT, Bard and, perhaps most intriguingly, law-specific platforms like Harvey, to provide substantive responses to legal questions is the most recent, but perhaps the most disruptive, to law practice. These tools require the user (the lawyer!) to input (“prompt” being the current AI lingo) the right legal questions, interpret the output from the tools, and apply the answers in the correct context—no easy task, but far easier than traversing to the law library or spending hours looking for the right provision. For instance, 50-state surveys may be a thing of the past. It is reasonable to think that in the coming three to five years, AI will be able to provide accurate, thoughtful output requiring little revision, all virtually instantaneously.
Sounds Great, But …
The greatest blessing of AI is, as many blessings are, a curse in some sense. AI will profoundly change the practice of law, and will likely do so within the next five to 10 years (at most). As such, there will be a sprint by firms and lawyers alike to embrace these technologies. We expect that practitioners that do not employ AI techniques and technologies will fall behind quickly, finding themselves unable to deliver the efficient, high-quality—and AI-supplemented—services clients will come to expect, and at the potentially diminishing costs they will demand. Our firm recognizes this and is already leveraging software in the above categories to help streamline processes and prepare more timely and better work product.
Importantly, we expect that this growth will need to be thoughtful and purposeful. First, while the outputs of AI will improve over time as the AI “learns,” lawyers will be fundamental to understanding those outputs, managing them efficiently and producing the ultimate work product, which not only can vary from client to client, but deal to deal. Second, lawyers that simply rely on AI as a crutch, to essentially replace the efforts of other lawyers, are prone to overreliance and lost critical thinking. Those lawyers will expose their practices to mistakes, to falling competitively behind lawyers that balance the use of AI with human experience and, worst of all, to malpractice and other ethical errors. There is no question that these are exciting times. Leaning in to AI and the promise it presents, learning about how to balance one’s own practice with AI’s assistance, and managing transactions using this unique, emerging technology will undoubtedly improve the practices (and lives!) of those lawyers that do embrace it. While we do not yet subscribe to the certainty of AI as a printing press-level of invention, the possibility cannot be discounted, which itself is remarkable.
This article may be reproduced, in whole or in part, with the prior permission of Thompson Hine LLP and acknowledgement of its source and copyright. It is intended to inform clients about legal matters of current interest and is not intended as legal advice. Readers should not act upon the information contained in it without professional counsel. This document may be considered attorney advertising in some jurisdictions. © 2023 THOMPSON HINE LLP. ALL RIGHTS RESERVED.