The Ai-intrigued Trial Lawyer

• Year: 2025
• Citation: Vol. 48 No. 2 Pg. 28
• Pages: 28

The AI-Intrigued Trial Lawyer No. Vol. 48 No. 2 Pg. 28 Wyoming Bar Journal April, 2025

I need to start with a confession: there are two types of AI relevant to this article, and as you will learn below, I am not an expert in either type. I generally understand that I have been using some form of Artificial Intelligence for years, that AI has many day-to-day applications, and that AI occasionally uses psychedelics and "hallucinates." I know lawyers have occasionally been burned when AI "trips"—and that scares me enough that I don't trust it to create legal arguments. Nevertheless, I remain cautiously intrigued enough to explore Al's imperfect playground. I have learned a few things along the way, some of which I share below.

My curiosity about using AI began in an effort to save time. I hoped that using a generative AI platform like ChatGPT would expedite my research making me a more productive and prepared lawyer. After all, AI is very good at creating summaries. I have used AI through ChatGPT to summarize the sometimes-dense scientific literature medical malpractice experts rely upon. This is particularly helpful when an expert cites numerous publications to support an opinion. You can also ask AI to search for articles or papers critical of the research an expert relied upon and ask it to identify any limitations in the research cited. This can provide ammunition to use during deposition or at trial. At a minimum, the information provides questions to ask your own expert during your preparation. A downside is that if you talk to your expert with AI-augmented knowledge, he or she may talk to you as if you know something, and that might leave you even more confused.

I have also used Adobe's AI Assistant to prepare summaries of cases in advance of oral arguments. Adobe's AI Assistant is good at summarizing cases if you specify the issues in the case that you are interested in. You can then combine these case summaries into a document and have the summaries at your fingertips when needed. Importantly, using AI to summarize cases is not a substitute for actually reading the cases. AI doesn't catch everything; it sometimes gets the holding wrong, and it does not know what issues are important to summarize unless you tell it what issues are important. If you tell AI to summarize the facts the court relied upon in deciding a particular issue, it will do a decent job, but it cannot determine what issues are actually important without your help.^[1]

AI is also very good at creating tables, generating lists, and preparing outlines from multiple sources. For example, AI can help create lists of topics to consider in evaluating a workplace

You can also upload multiple documents into ChatGPT and ask it to create an outline by topic and then a slide show for presentation, or a comparison chart.

accident involving a mobile crane and in preparing a list of items for a preservation letter in a case involving chemical exposure in a laboratory. In the former, AI might rely upon applicable and generally available federal regulations and ANSI standards, and in the latter, AI might use protocols you've previously obtained to analyze recordkeeping requirements and documentation protocols for a particular lab. This is easy to do by uploading a document into ChatGPT.^[2] You can also upload multiple documents into ChatGPT and ask it to create an outline by topic and then a slide show for presentation, or a comparison chart.

AI is also particularly good at creating code, although this is not something I hope to have to routinely use. Recently, I wanted to create an exhibit showing changing lab values for my client over time. After providing ChatGPT with all of the data, it indicated that it could not create the graph I wanted, but that it could send me Python code so I could create the graph in a different...