The use of artificial intelligence in legal cases has come under closer examination after a recent lawsuit involving a deadly factory explosion in Texas. Plaintiffs in the case, which resulted in three fatalities and extensive property damage, alleged that an expert witness used ChatGPT, an AI chatbot, to generate more than 80% of his expert report. The lawsuit was filed against 3M, a manufacturer, following a January 2020 incident at one of their facilities. The expert, Jason Autenrieth, was hired by 3M just before the trial to evaluate whether the company met the required safety standards and whether its technicians performed necessary checks. During the trial, plaintiffs' attorneys questioned whether the report provided by Autenrieth was entirely his own work. A document called a citation overlay, which listed the sources used in the draft report, revealed the extent of AI involvement. One of the plaintiffs' attorneys, Adam Lewis, expressed surprise at how extensively Autenrieth used ChatGPT, including sending 300 pages of prompts to the AI. These prompts included inquiries about specific machinery and attempts to challenge the opposing side’s witness. Autenrieth initially denied that the AI wrote his report but later admitted it helped with formatting and other tasks. After the deposition, 3M removed him from their list of witnesses, but the plaintiffs obtained a court order requiring him to testify about the prompts he used with ChatGPT. During the trial, Autenrieth defended his use of AI, stating he believed it was not wrong because he already had his own opinion on the case. In a statement, 3M emphasized that Autenrieth was an independent third-party expert, not an employee, and that he used AI as a tool to assist in preparing his report. Autenrieth’s lawyer, Janina Wojtkowski, added that he did not rely on AI to form his opinions and that his professional judgment remained intact. The jury ultimately found 3M negligent in its safety checks, and last month, the court awarded victims $61.5 million. A 3M spokesperson said they disagree with the verdict and plan to appeal. Legal experts have raised concerns about the potential risks of using AI in legal proceedings. Maura Grossman, a professor at the University of Waterloo, warned that AI tools may not provide objective or neutral responses and advised legal professionals to be cautious when using them. She emphasized the need for due diligence, including disclosing AI use and verifying the accuracy of AI-generated content. Damien Charlotin, an adjunct professor at the University of California College of the Law, noted that courts are increasingly encountering issues with AI, such as false information and incorrect citations. He compiled a database showing over 1,300 instances of AI-related errors in U.S. legal cases since 2023.