Imagine navigating a federal civil case without a lawyer while confronting extensive procedural rules, strict deadlines, and complex filing requirements. For a growing number of individuals, accessing the civil justice system means doing so without legal representation.
Research by Dr. Or Cohen-Sasson, Lecturer in Law, Law & Technology Fellow, and director of the Miami Law & AI Lab at the University of Miami School of Law, examines this access-to-justice challenge. His latest study on artificial intelligence and civil justice was accepted for presentation at AI4Law, the legal workshop held in conjunction with the 2026 International Conference on Machine Learning in Seoul.
Because law schools rarely present research at ICML, the study represents a notable connection between legal scholarship and computer science.
Cohen-Sasson conducted a large-scale, data-driven investigation of 2.8 million federal civil cases filed between 2008 and 2025, along with nearly 13,000 complaints filed by individuals without attorneys. His findings document a significant shift in how people interact with the federal civil justice system.
During the study period, the proportion of individuals representing themselves in federal civil cases, known as pro se litigants, more than doubled, increasing from approximately 11% to more than 22%. In other words, more than one in five federal civil cases now involves a person proceeding without legal counsel.
Navigating a court case without an attorney can increase the risk of procedural errors and dismissed claims and can place additional demands on courts. Artificial intelligence tools have been proposed as one way to help self-represented individuals prepare documents and understand legal processes, but technology developers need reliable data to design tools that address users’ needs effectively.
