AI + Civil Litigation

"AI engines are designed to keep users coming back. In contrast, the law is a consistent set of standards that are not designed to make everyone happy, or to feel any emotion." Torres v. Spraker, Case No. EF2025-12556, 2026 WL 1450457, *4 (N.Y. Sup. Ct. May 20, 2026).

This judge spent some time addressing the differences between the "sycophancy or flattery" of AI machines and their outputs against "blind justice" and human, legal reasoning. Id.

My next favorite bit? "The consistent application of the law is why litigants can leave a court proceeding feeling upset or bewildered. Court is not a feedback loop meant to please everyone . . ." Id. Preach.

New York continues to crank out AI-hallucination cases. This one went further than most recent cases, highlighting the need for and importance of human and legal reasoning that is "beyond the capacity of a computer search" or AI machine. Id.

This issue will persist--for pro se litigants and poorly-trained attorneys--and these frank reminders will continue to ring true along the way:

"If the same information is typed into a chatbot, however, the result is tainted by the unfortunate feedback loops built into the system to reinforce one's misguided or misunderstood conclusion and inferences. The Court is not an AI chatbot, it is an established institution of government." Id.

"An AI application is incapable of reasoned legal conclusions as it has no reasoning capability, it is solely generating results from a prompt and searching for patterns in its datasets to generate a response that is engaging to the user, not necessarily accurate. AI cannot be the sole source of legal determinations as the application of the statutes, rules, and caselaw requires the unique reasoning ability of the human mind, as well as respect for the determinations of existing, past legal disputes and interpretations." Id. at 6.

♥️✌🏻🔥