ABSTRACT
The United States has long suffered from unequal access to justice, with countless low-and middle-income Americans forced to navigate the legal system alone. Recently, prominent judges, lawyers, and scholars have seized on generative AI as a potential corrective. These techno-optimists maintain that ChatGPT and other large language models can demystify the law and address unmet legal needs. Chief Justice John Roberts has proclaimed that AI-based tools ‘have the welcome potential to smooth out any mismatch between available resources and urgent needs in our court system’.
Although AI will be an increasingly important source of legal assistance for underserved populations, the techno-optimists’ view is oversimplistic in several respects. Asymmetric information hinders underserved communities’ ability to use AI effectively, exposing them to hallucinations and misinformation. Legal needs could also grow exponentially as the wealthy and powerful deploy AI against low- and middle-income people in situations where hiring lawyers is currently impractical. As with previous technological innovations, AI’s impact must be assessed in the context of a deeply unequal and adversarial justice system.
This Article proposes two key reforms for integrating AI into the civil justice system. First, it calls for the training of justice tech workers to counsel unrepresented individuals in the responsible use of AI-based legal tools. Second, courts should bolster existing ethical requirements and mandate the reasonable verification of factual claims and legal authorities in proceedings that commonly involve unrepresented parties. Without these reforms, AI will only entrench and amplify longstanding inequalities in the justice system.
Markovic, Milan, Equal Justice amd Generative AI (October 14, 2025), 87 Ohio State Law Journal (2026).
Leave a Reply