Donald Sauve’s handwritten complaint got tossed from federal court quickly. When he returned with AI-assisted filings—polished, numerous, and properly formatted—clerks faced hours of extra work captioning and entering each document. Sauve represents a growing wave of self-represented litigants using chatbots to flood court systems with paperwork that looks professional but often lacks substance.
The Numbers Tell the Story
Recent MIT and USC research analyzing 4.5 million federal civil cases reveals the scope. Self-represented litigation jumped from a steady 11% average to 16.8% by fiscal 2025. More concerning: docket entries from these cases spiked 158% above pre-AI averages. This means each case generates far more paperwork than before.
By 2026, researchers flagged over 18% of complaints as likely containing AI-generated text. Your local courthouse isn’t just seeing more cases—it’s drowning in documentation that someone’s algorithm helped craft.
Access Versus Abuse
The technology cuts both ways. Legal Services Corporation groups legitimately use AI for high-frequency matters like workers’ compensation and landlord-tenant disputes. This helps people who can’t afford lawyers navigate complex procedures.
Judge Michael Y. Scudder of the Seventh Circuit acknowledges AI’s “great promise” for improving access to justice. Yet courts already struggle with AI problems from actual attorneys—fabricated citations, hallucinated cases, outright fiction presented as precedent. Now multiply that risk across thousands of self-represented litigants with zero legal training but unlimited access to persuasive ChatGPT.
System Strain and Solutions
Every AI-generated filing still requires human review. Clerks must read, caption, and enter each document regardless of quality. The administrative burden multiplies when a single case produces dozens of polished but potentially meritless motions.
The Seventh Circuit recently warned that accuracy and honesty remain paramount in AI-assisted filings. Self-represented litigants can’t blame the bot for false claims. Courts are responding with disclosure rules, sanctions for AI-generated lies, and stronger screening procedures. The future likely holds court-approved legal chatbots designed for legitimate self-help rather than open-ended document generation.
The tension persists: democratizing legal access while preventing system abuse. Courts must adapt to this new reality where anyone can produce lawyer-quality paperwork—whether their underlying claims deserve it or not.
