AI Judging: Dystopian Fantasy or New Reality? Results of Using AI to Decide 1,000+ Federal Appeals

Can AI perform judicial work traditionally entrusted to judges and law clerks? Professor Nick Datzov put the question to the test with more than 1,000 actual federal appeals. A professor at the University of North Dakota School of Law, Nick’s also a nationally recognized expert in artificial intelligence law, award-winning teacher, and popular CLE presenter.

In this presentation, he’ll share findings from his first-of-its-kind empirical study, co-authored with David Schwartz (Northwestern Pritzker School of Law), comparing the work of generative AI and the U.S. Court of Appeals for the Eighth Circuit over two years. Applying his prior AI Judge taxonomy to a novel dataset of federal appeals, the study tested AI’s ability to assess case complexity, categorize cases, and even decide the appeals themselves.

You’ll learn:

• How often AI reached the same result as the Eighth Circuit
• Which types of cases AI was best equipped to handle
• Where AI performed well and where it struggled
• Whether appellate attorneys should start consulting AI regarding their chances on appeal

Along the way, Nick will also explore what the results reveal about the possibilities – and limits – of AI-assisted judging. His work is a data-driven look at a question that has made headlines and is quickly moving from theoretical to practical: What role can AI actually play in the administration of justice?

– Nikola L. Datzov; University of North Dakota School of Law; Grand Forks, North Dakota

Attendees of the 2026 Midwest IP Institute (in-person or replay) may attend this webcast for FREE.
Instructions on how to register for free will be distributed to attendees at the Institute.