This is the second installment of a blog series exploring the profound ways Artificial Intelligence (AI) is transforming the landscape of litigation, presenting both advancements and significant challenges. The article specifically delves into the Mendones v. Cushman and Wakefield, Inc. case, a critical example where litigants attempted to introduce AI-generated and substantially altered exhibits as legitimate evidence. This case underscores the inherent risks associated with relying on AI in legal proceedings and emphasizes the paramount importance for attorneys and litigants to remain acutely aware and diligent concerning AI's evolving influence throughout the entire litigation process, highlighting potential interferences with justice.
Background
The article details the California state court case, Mendones v. Cushman and Wakefield, Inc., where plaintiffs Ariel and Maridol Mendones were accused of deliberately submitting false testimony, including nine AI-generated or altered exhibits, to support their motion for summary judgment. A key example cited by the court was a video testimony purportedly from Geri Haas, which the court found to be generative AI, noting significant discrepancies in accent, cadence, volume, word choice, pauses, gestures, and facial expressions compared to authentic testimony. The AI-generated "person" displayed robotic characteristics and a mismatch between mouth movements and spoken words. Additionally, fabricated text messages, whose veracity "strained credulity," were submitted. The court acknowledged its limitations, stating it lacked the "time, funding, or technical expertise" for a comprehensive forensic analysis of all suspicious submissions. After considering monetary penalties and criminal referral, the court ultimately imposed a terminating sanction, dismissing the lawsuit. This severe penalty was deemed "proportional to the harm" caused by the plaintiffs' misuse of court processes, sending a clear message of zero tolerance for AI "evidence."
Not-So-Futuristic Applications
This section reflects on the Mendones case as a prophetic fulfillment of prior discussions regarding the potential for AI-generated witness testimony in court. The article notes that while current AI-generated content might still be detectable by a discerning eye, its mere introduction into active litigation is deeply troubling. A significant concern raised is that many courts, grappling with increasingly burdened dockets, tight timelines, and numerous deadlines, may not possess the necessary resources, time, or technical expertise to thoroughly analyze and detect all instances of fabricated or manipulated evidence, as the Mendones court itself admitted. As AI technologies continue to advance rapidly in sophistication, the inherent risk of such fabricated evidence not only being submitted but potentially being admitted as legitimate material heightens considerably. The authors argue that this capability to create non-existent evidence crosses a critical line from beneficial "artificial intelligence" into "artificial interference," fundamentally jeopardizing an opposing party's constitutional right to a fair trial by undermining the very foundation of evidentiary integrity.
Key Takeaways for Litigants
This section provides crucial advice for individuals involved in legal disputes in the age of AI. Firstly, it strongly cautions that using AI in litigation comes with substantial risks, particularly when used for fabricating evidence. While AI can potentially streamline legal tasks like research or data synthesis, its misuse for creating false evidence can result in far more severe consequences than any attorney's fees saved, including the ultimate penalty of case dismissal, as seen in Mendones. Secondly, represented litigants are urged to proactively request that their attorneys disclose any use of AI tools in their legal work, ensuring they are fully informed of potential risks and ethical implications. Thirdly, litigants must be fully prepared to confront and meticulously vet any AI "evidence" introduced by opposing parties at any stage of the litigation process, from initial pre-suit demands through discovery and trial. The article suggests comprehensive inquiries into the method and manner of the evidence's creation, the individuals involved in its generation or manipulation, details about prompt changes leading to the final output, and whether the evidence has undergone review and approval by qualified third-party experts. The core message is clear: if evidence appears "too good or too bad to be true," it warrants extreme scrutiny, as it might indeed be artificial. Ensuring complete transparency and authenticity of all evidence is vital to uphold the integrity of the justice system and secure a fair proceeding.
