Preprint finds LLMs driving pro se federal court filings surge after ChatGPT release
Researchers from MIT and USC analyzed 4.5 million U.S. federal civil court cases from 2005–2026, finding pro se filings rising from 11% to 16.8% in FY2025, concentrated in formulaic categories coinciding with ChatGPT’s 2022 release. Pro se cases had 158% more early docket entries than represented cases. Pangram AI detector flagged AI-generated content in 8.2% of post-AI pro se filings versus 0.1% pre-AI, with detection spiking after ChatGPT launch.
Every system that was regulated, either explicitly or implicitly, by the fact that they were effortful for humans (letters of recommendation, lawsuits, government filings, essays) will break.
🚨New preprint! We find evidence of LLMs enabling people to file lawsuits without lawyers (filing "pro se") at historically unprecedented rates in federal courts.👇 1/n
Wrote about this two years ago (and expanded on this in my book): https://www.oneusefulthing.org/p/setting-time-on-fire-and-the-temptation
Every system that was regulated, either explicitly or implicitly, by the fact that they were effortful for humans (letters of recommendation, lawsuits, government filings, essays) will break.
AI is good for society, Exhibit N.
🚨New preprint! We find evidence of LLMs enabling people to file lawsuits without lawyers (filing "pro se") at historically unprecedented rates in federal courts.👇 1/n