Chapel Hill, N.C. – The University of North Carolina School of Law staged a novel mock Supreme Court hearing on Tuesday, featuring artificial‑intelligence justices powered by ChatGPT. The event, titled “May It Please the Machine,” gave students the chance to argue a real Supreme Court case before a panel of AI judges.
How the AI justices were set up
Faculty members uploaded the case’s briefing documents and created detailed “persona files” for each AI justice, assigning them distinct legal perspectives. The AI was isolated from the internet and any external sources, ensuring it could only rely on the supplied materials. During the hearing, the AI justices listened to the arguments, asked follow‑up questions, and referenced each other’s remarks, mimicking the dynamics of a true Supreme Court conference.
Dean Downing’s goals and assessment
Nicole Downing, UNC School of Law’s dean for artificial‑intelligence initiatives, emphasized that the experiment was not intended to replace judges. “It can apply law, it can do some reasoning,” she said, “but the attorney is still going to be so valuable to that process. There is so much originality, creativity, that comes from a human.” The AI panel ultimately reached the same decision as the actual Supreme Court in the case, but Downing noted significant room for improvement, especially in conversational fluency. She rated the AI justices a modest 1.5 out of 10 for their overall performance.
Implications for legal practice
The exercise reflects a growing interest among law schools nationwide in exploring how AI tools can augment legal research and argumentation. By testing AI’s capacity to parse complex legal briefs and engage in dialogue, UNC hopes to better understand the technology’s strengths and limitations. The school plans to refine the AI’s training data and interaction models before expanding the program to other courses.
Student reactions
Participating students described the experience as both challenging and enlightening. One student noted that the AI’s questions forced them to clarify points they might otherwise have taken for granted, while another remarked that the limited conversational ability highlighted the importance of human intuition and advocacy.
Future directions
Downing indicated that future iterations will incorporate more sophisticated language models and broader data sets, aiming to improve the AI’s ability to reason about nuanced legal principles. The school also intends to publish findings on how AI assistance impacts student learning outcomes, hoping to contribute to the national conversation on responsible AI integration in the legal profession.
As AI continues to reshape many professional fields, UNC’s experiment underscores a cautious optimism: technology can serve as a powerful tool, but it cannot replace the moral judgment, creativity, and personal touch that human lawyers bring to the courtroom.
Original reporting: WRAL Raleigh — read the source article.
