U.S. Judge: Lawyers' reliance on AI is a "dangerous shortcut"

Specifically, in the context of a labor dispute between Ms. Cynthia White (plaintiff) and Walmart (defendant – one of the largest retail corporations in the United States), attorney Mark Waterfill, representing the plaintiff, raised objections to Walmart’s responses concerning requests for the production of evidence in the case. However, according to the assessment of Judge Tim Baker, these objections were not directly researched and drafted by the attorney himself, but were instead generated by an AI software. According to the case file, attorney Waterfill uploaded Walmart’s responses to an AI software and requested the system to identify deficiencies in the evidence production process. He then copied the results generated by the AI software and used this content to communicate with counsel for the defendant, while also incorporating such content into documents submitted to the court. The AI software used by Waterfill was not specifically identified.

At the hearing, the attorney admitted to using AI software in the manner described above. Judge Baker held that attorney Waterfill had violated the obligation to personally review and assess the deficiencies in the discovery process identified by the AI software before using such content to object to Walmart. This also served as the basis for the court’s determination that the attorney had failed to fulfill professional responsibilities by delegating content verification to artificial intelligence.

From the developments of the case, it can be seen that the issue does not lie with artificial intelligence itself, but rather with how it is used by humans. Ultimately, AI remains merely a data-processing system that may produce errors or inaccurate information if not properly verified. Therefore, the use of AI cannot be equated with relinquishing the responsibility of content verification. In litigation practice, attorneys are required to directly review, assess, and assume responsibility for the content submitted to the court. The use of supporting tools, including AI software, does not alter this obligation.

In the case between Cynthia White and Walmart mentioned above, the attorney’s uncontrolled use of artificial intelligence in drafting litigation documents reduced both the quality of the content and the accuracy of the information. This not only affects the attorney’s professional credibility, but may also prolong the resolution of the case and increase litigation costs.

At a more serious level, if inaccurate content is not detected in a timely manner, it may influence the court’s assessment of the case. In this context, the risk of misuse of AI software in litigation, whether intentional or unintentional, raises the need for stricter control over the use of such tools.