AI Chatbots Are Flooding Courts With Self-Represented Lawsuits

Federal courts across the United States are experiencing an unprecedented surge in lawsuits filed by people without lawyers, and judges are pointing to artificial intelligence as the primary driver.

According to a study examining 4.5 million federal civil cases from 2005 to 2026, the share of self-represented lawsuits increased from 11% in 2022 to 16.8% in 2025. The number of individual filings more than doubled from pre-2023 levels, as first reported by MIT Technology Review.

Researchers Anand Shah at MIT and Joshua Levy at the University of Southern California tested 1,600 randomly sampled court documents using Pangram, a commercial AI-text detector. The share flagged as containing AI-generated writing rose from just 1% in 2023 to 18% in 2026.

Why it matters

This shift represents a fundamental change in access to the legal system, with AI tools potentially democratizing legal representation while simultaneously raising urgent questions about liability, legal ethics, and the quality of justice. Courts and lawmakers are now grappling with whether chatbots should be treated like lawyers—with similar duties, privileges, and accountability—even as people increasingly depend on them for legal guidance.

Judges See Better Documents, Not Better Outcomes

Judge Maritza Braswell, a federal magistrate judge in Colorado, says she can recognize AI-generated legal writing by its prose style and occasional hallucinated cases or fabricated quotes. Despite these flaws, she finds AI-assisted filings easier to process than the handwritten documents that previously dominated pro se cases.

"I have to be really careful because some of them contain hallucinations and errors, but I can generally understand what they're arguing better with AI assistance from them than without it," Braswell told MIT Technology Review.

However, the clearer writing hasn't translated into better results. The study found that self-represented litigants remain far more likely to lose their cases than those with human lawyers, even with AI assistance.

"It turns out that mounting a lawsuit is a complex, multifaceted task. Not all of it is just drafting text," Levy noted.

Legal Gray Areas Emerge

Courts are now confronting novel questions about AI's role in legal proceedings. In February, a Michigan federal court ruled that a self-represented person's conversations with ChatGPT qualified as protected work product. That same day, a New York federal court reached the opposite conclusion regarding Claude, arguing that AI companies can disclose user data to third parties, eliminating expectations of confidentiality.

Judge Allison Goddard in California has observed ChatGPT providing wildly inaccurate case valuations during settlement negotiations, leading one slip-and-fall plaintiff to demand $700,000 for a case worth far less.

The liability question came to a head in March when Nippon Life Insurance Company sued OpenAI, alleging ChatGPT practiced law without a license. OpenAI responded in May by arguing that "ChatGPT is not a person and neither has nor uses any degree of legal knowledge or skill." The case remains pending.

Legislative Response Begins

New York introduced legislation in March that would bar chatbots from impersonating lawyers, even with disclosure notices. Federal bills have been proposed to ban chatbots from posing as licensed professionals, though none have gained significant traction.

Online communities are accelerating the trend. A viral Reddit post in December 2024 walked immigration applicants through using Microsoft Copilot to draft legal documents, contributing to a surge in Vermont filings from 45 annually before 2022 to more than 1,100 in 2024.

These findings were detailed in a comprehensive report by MIT Technology Review.