Pro se - meaning "for oneself" in Latin - refers to people who argue their own legal cases without hiring an attorney. Courts have always accommodated self-represented litigants, but filing a lawsuit was naturally limited by how much effort it required. Drafting a complaint, citing relevant precedent, formatting motions correctly - all of it takes time and legal knowledge most people don't have.
AI tools have removed most of that friction, and courts are now seeing the results.
ChatGPT, Claude, and similar tools can produce formal-looking legal documents in minutes. A non-lawyer describes their situation in plain language and gets back a structured complaint with headings, statute citations, and arguments formatted the way law school trains attorneys to write. For someone with a genuine grievance who can't afford $300-per-hour attorney fees, this is useful.
The problem is what those documents often contain.
When AI Gets the Law Wrong
Legal filings aren't just polished writing - they're factual claims about what the law says. AI tools have a documented tendency to hallucinate case citations: producing references like "Jones v. Smith, 2019, 9th Circuit" that look real but don't exist. When a judge or opposing counsel tries to look up the case, there's nothing there.
This isn't a theoretical risk. Judges in multiple jurisdictions have sanctioned licensed attorneys for submitting AI-generated briefs with invented case law. Courts are now seeing the same problem from pro se filers who didn't know to verify every citation the AI produced.
Beyond fabricated precedents, there's a volume problem. Civil courts were already under-resourced before AI-assisted filing became practical. When the labor cost of preparing a lawsuit drops to nearly zero, every grievance becomes a candidate for litigation in a way it wasn't before.
Courts Are Starting to Push Back
Some federal courts have issued standing orders requiring disclosure of any AI use in legal filings. A handful of judges now explicitly require verification that cited cases exist. Enforcement is difficult - there's no reliable technical method to detect AI-generated text, and courts don't have capacity to audit filings before docketing them.
For attorneys on the other side, AI-assisted pro se filings create real headaches: they have to verify opposing citations that no attorney would have fabricated, and briefing schedules stretch when waves of AI-drafted motions arrive.
The deeper issue isn't that AI is breaking the legal system. Access to courts has always been rationed by cost, and AI tools are disrupting that arrangement fast. Some filings flooding courthouses come from people with legitimate grievances who couldn't previously navigate the system alone. Others are speculative or built on AI-invented facts. Courts weren't designed to sort through high volumes of AI-assisted filings quickly, and the procedural rules built to manage self-represented litigants were written for a much smaller population than AI tools are now creating.