What is common knowledge? Consider consulting ChatGPT, as the panel in Ross v. United States, a case before the District of Columbia Court of Appeals did. In an opinion issued on February 20, 2025, the majority and dissenting judges publicly disclosed and discussed their use of artificial intelligence (“AI”) tools—specifically, ChatGPT—to assist in reaching a determination regarding whether the government’s reliance on “common knowledge” was sufficient to fill critical gaps in the government’s case. Following in the footsteps of a federal appellate court, dissenting Associate Judge Deahl used ChatGPT to make a point regarding common knowledge, while the majority responded with its own ChatGPT example. For law firms, clients, and legal practitioners, this opinion signals the judiciary’s growing interest, reliance and use of AI tools in the courtroom.
In the case, the government charged Niya Ross with one count of cruelty to animals under D.C. Code §§ 22-1001, 22-1002 for leaving her dog, Cinnamon, in a car on a hot day in September 2023. During a bench trial, the trial court found Ross guilty. On appeal, Ross challenged the decision on two grounds. First, she argued that the government failed to demonstrate that her dog suffered from being left in the car, particularly because the government did not introduce expert testimony. Second, Ross argued that the government had not established that she possessed the requisite intent under the criminal statute: general intent with malice.
What sets this case apart is the transparency regarding the use of AI in the judicial process. Associate Judge Deahl, in his dissent, explicitly stated that he consulted ChatGPT, an AI-powered large language model developed by OpenAI, to explore what he described as “common knowledge” about the effects of leaving a dog in a hot car. He posed hypothetical questions to ChatGPT, such as whether it is harmful to leave a dog in a car with the windows cracked for an hour and twenty minutes at 98 degrees Fahrenheit, and referenced its responses to bolster his reasoning.
The majority, in turn, responded by also engaging with ChatGPT, using it to analyze a hypothetical question related to a prior case, Long v. United States (156 A.3d 698), to contrast the dissent’s approach and underscore the limitations of relying on AI for factual or legal conclusions. Associate Judge Howard, in a concurring opinion, further elaborated on the broader implications of AI in judicial decision-making, drawing on his involvement with the D.C. Courts AI Task Force and addressing ethical, security, and privacy concerns.
This open acknowledgment is a landmark development, as few, if any, prior published judicial opinions in the U.S. have detailed the use of AI tools in judicial opinions with such specificity. The court’s discussion builds on Judge Kevin Newsom’s concurring opinion in Snell v. United Specialty Ins. Co. (102 F.4th 1208, 11th Cir. 2024), which explored AI’s role in judicial reasoning, but Ross takes the conversation further by integrating AI directly into the opinion-drafting process.
Ross v. United States is a pioneering decision and the start of what is likely to become a trend of judges (or their clerks) trying out AI tools. As the legal profession navigates the intersection of technology and law, this opinion serves as both an opportunity and a cautionary tale. Law firms and clients should stay informed about the evolving use of AI in courts and the legal profession, assess its potential benefits for legal practice, and address the associated risks to maintain ethical and effective representation.
For assistance with IP or AI questions, please contact Victor Johnson, Sara Gates, or anyone from the Dentons US Intellectual Property & Technology Group.