Imagine sitting in court, getting ready to hear a victim impact statement during a sentencing hearing, but instead of hearing a family member deliver the impact statement about the decedent, you see a video. Not just any video, but an AI generated video of the deceased themselves. A video where the victim has been brought back to life to deliver an emotional impact statement. This is no longer a scenario from a sci-fi movie or from imagination.
This is where litigation and trials are now. The use of AI is rapidly embedding itself into the legal process, and shows no signs of stopping. Recently in Arizona, Stacey Wales created an AI video of her brother, who was tragically killed in a road rage accident, delivering his victim impact statement at his killer’s sentencing hearing. While this was a creative, and rather harmless use of AI in the legal, there have been more sinister uses of AI in the courtroom. Jerome Dewald, representing himself in an employment dispute in front of the New York State Supreme Court Appellate Division’s First Judicial Department, used an AI generated “attorney” to deliver his oral arguments. Justice Sallie Manzanet-Daniels almost immediately halted the presentation when it dawned on her that the person on the screen was not real.
While those are the two newsworthy moments of AI use in the courtroom in 2025, it is a guarantee that more are to come. Additionally, while these two incidents were done by non-attorneys, if this pattern continues, there is no doubt it will start becoming the norm for attorneys to use AI in similar manners.
As of now, outdated guidelines exist for the use of AI in the courtroom; these two incidents in 2025 alone are sufficient to demonstrate why we need stronger, more robust universal rules regarding the use of AI in the litigation process. Furthermore, a large number of the rules revolve around the disclosure of the use of AI, not the use of AI itself. Since the Federal District Court for the Northern District of Texas, namely Judge Brantley Starr, issued a standing order regarding AI disclosures in 2023, several other district courts have followed suit. However, a majority of these standing orders were drafted with the idea of preventing AI hallucinations in court filings and motions. Now in 2025, hallucinations have decreased, making it necessary to create a universal approach and guideline to how we approach AI use in the litigation process.