I find the endless stream of updates and new releases about AI overwhelming; trying to stay informed feels like to perpetual homework. And yet there seems to be a massive gap between the buzz on LinkedIn, news articles etc, and its practical applications in our daily work. A friend of mine, who is a lawyer, mentioned that she doesn't understand the hype surrounding AI in law because it hasn't altered her work or personal life at all. She asked me directly, "What are lawyers actually using it for right now?" Here are 3 of the most common used cases for generative AI with examples (illustrative not comprehensive) of their application in different practice areas of law.
Summarization: Lawyers frequently turn to AI for creating succinct summaries or detailed timelines of events.
Litigation: summaries are required communications with opposing counsel, instructions to barristers, affidavit drafts, and are the background of the case in pleadings or submissions to the WRC.
Immigration law: summaries are part of cover letters to embassies, the Department of Justice, as well as in case submissions.
AI's summarizing capabilities are a subset of its broader application in legal drafting, which include more complex tasks such as drafting commercial contracts.
Extracting information from multiple sources: This is generally the first part of the workflow for lawyers cross diverse practice areas.
Litigation: Lawyers gather and analyze a variety of information, including personal data, party correspondence, contracts, medical and expert reports, and witness accounts. These details are critical for completing procedural forms, such as the Personal Injuries Assessment Board (PIAB) form and drafting pleadings. Gen Ai can also identify what information is missing and pinpoint inconsistencies which is invaluable in verifying the integrity of witness statements.
Corporate Law: The focus is on compiling shareholder documents, agreements, and memoranda to facilitate company registration, draft contracts, and prepare board minutes and ensuring corporate agreements reflect the initial terms agreed upon.
Immigration: Essential documentation includes passports, visas, financial statements, employment contracts, and academic records, all of which are necessary to fill out immigration forms accurately and gen AI can once again, advise on what documentation still needs to be furnished by the applicant.
Brainstorming: For me this is the most powerful used case
Litigation: Generative AI serves as a brainstorming partner, even for seasoned lawyers, by suggesting case strategies, counterarguments that the opposition might raise and defences to those arguments. It can also creatively identify mutual goals for settlements or mediation agreements.
Commercial Agreements: Beyond identifying high-risk clauses for clients and propose alternative language for drafting agreements, when facing a particularly contentious counterpart, AI can offer innovative solutions that think outside the box and go beyond standard contract language to effectively mitigate risks. For example, I was drafting an IP agreement and it suggested having an escrow arrangement for the source code and establishing a joint IP monitoring committee.
The practicality of generative AI in legal practices is intrinsically tied to the digitalization of workflows, the quality of digital inputs, and the complexity of the output required. While AI offers significant benefits, such as drafting initial affidavits in litigation, its effectiveness is curtailed in areas like conveyancing, where processes remain paper-based and inputs like deeds and certificates are not consistently digitized. The transition to digital relies not on lawyers but on institutions like the Irish courts and the PRA, which are currently navigating their own paths towards digitalization.
Finally, we must recognize the limitations of standard generative AI. While it excels at generating text, it falls short in performing calculations. For instance, it's pretty good at the text for an affidavit from notes and multiple text documents, but it struggles with tasks like analyzing bank statements to categorize monthly income and expenses for an affidavit of means. However, there are specialized AI tools out there designed for financial analysis. I haven't had the chance to try it yet, so it seems I've got another item to add to my growing list of things to check out.