AI-Written Lawsuits Are Flooding US Courts and Judges Cannot Keep Up

Artificial intelligence tools are making it easier than ever for people to file lawsuits without a lawyer, and the results are overwhelming the American court system. A new study from researchers at MIT and USC reveals just how dramatic the shift has become.

A Flood of Self-Filed Cases

The study, led by MIT researcher Anand Shah and USC's Joshua Levy, analyzed 4.5 million federal civil cases and 46 million court records. The findings are striking. Self-filed cases from non-incarcerated individuals, known as pro se filings, jumped from a long-standing 11 percent of all federal civil cases to 16.8 percent by fiscal year 2025. That might sound modest, but the activity within those cases surged by 158 percent in the first 180 days compared to the period before AI tools became widely available.

Even more revealing, the researchers sampled 1,600 federal complaints and found that 18 percent of filings in 2026 contain AI-generated text. Before the rise of large language models, or LLMs, which are the technology behind tools like ChatGPT, that number was essentially zero.

Promise and Peril

The implications cut both ways. On one hand, AI is genuinely democratizing access to the legal system. People who cannot afford attorneys, sometimes called self-represented litigants, can now use chatbots to draft legal documents, understand procedures, and file complaints. That is a meaningful step toward equal justice.

On the other hand, courts are struggling to keep up. Shah warned that if the trend continues, courts could "basically have to grind to a halt." The problem is not just volume. AI-generated filings sometimes contain fabricated legal citations, a well-documented issue known as hallucination, where the model invents case law that does not exist. In April 2026, even the prestigious law firm Sullivan and Cromwell apologized after AI-generated fake legal material appeared in a bankruptcy matter.

The research points to a simple economic principle that applies to every industry AI touches: when you lower the cost of entry, demand increases. For the legal system, that means more filings, more work for judges, and a growing need for new tools to manage the AI-driven caseload.