AI in Legal Proceedings Fuels Sharp Rise in Pro Se Litigants, Straining US Courts

No lawyer? No money? More Americans are suing with AI help

May 15 (Reuters) - Las Vegas resident Nicole Silverberg says she called 20 lawyers in hopes of finding one willing to sue her former landlord on a contingency basis because she couldn't afford to pay legal fees upfront. When none agreed, she turned to artificial intelligence. Using the paid versions of OpenAI's ChatGPT and Grok, she filed a federal lawsuit last summer against the landlord, alleging she was evicted in retaliation for withholding rent because of ongoing water contamination issues. The tools guided her on everything from how to file a preliminary injunction motion to how to hold a defendant in default. Silverberg said the technology helped her push the case further than she could have on her own, including helping spot errors and weaknesses from opposing counsel, despite occasional hiccups. "I didn't know AI could be wrong at first," she said. "It was like telling me Santa Claus was not real." Silverberg is part of a wave of pro se litigants using AI tools to be their own advocate in court. New research, opens new tab from two economics Ph.D students shows that self-represented litigants account for nearly 17% of all federal civil court cases in the fiscal year that ended September 30, 2025, up from a long-term average of about 11%. The federal appellate court system separately reported a 9% increase in appeals by pro se litigants in fiscal 2025. AI is helping fill long-standing gaps in legal-aid services, which leave many would-be plaintiffs on the sidelines if they don't have the means or wherewithal to hire an attorney. Unlike in criminal cases, where low-income defendants are assigned a public defender, those in civil disputes must pay their own way. Government-funded legal-aid organizations and nonprofits ⁠help fill the void, but funding fluctuates and has been in decline. The Legal Services Corporation estimated in a 2022 report that low-income Americans don't get any or enough legal help for 92% of civil legal problems. Anthropic this week announced several partnerships that it said will help get more AI legal tools into the hands of people and small businesses that can't afford traditional legal services, as part of a broader rollout of Claude's AI features for lawyers. It's too early to tell whether the increase in AI use is actually improving outcomes for pro se plaintiffs, but several people I spoke to said they find it encouraging that more individuals are turning to the courts. AI is enabling more filings "that at least get past the clerk's gate," said Sonja Ebron, the CEO of Courtroom5, an AI-powered paid online service that helps pro se litigants. Some Courtroom5 tools are now available for Claude users following this week's announcement. Even as technology becomes more sophisticated, using ChatGPT or other AI to stand in for a lawyer has major risks. For every smartly worded pro se filing powered by AI, there are others filled with hallucinated case law and excess pages. The increase in cases is also stressing already burdened courts and risks extending the time it takes for cases to resolve. That tension between access to justice and strain on the system is one that judges, attorneys and legal-industry watchers are thinking about. "The courts will probably struggle to keep up if trends continue as they are," said Joshua Levy, an economics Ph.D. student at the University of Southern California, who co-authored the new paper showing the rise in pro se litigants. In addition to the rise ⁠in cases, Levy and co-author Anand Shah, a doctoral student in economics at the Massachusetts Institute of Technology, found the dockets in pro se cases getting heavier, with more entries, more briefs and more arguments. Levy and Shah found the number of docket entries per case up 38% in mid-2025 compared to the pre-AI mean. They caveat in the paper, which they're seeking to get published, that they can't prove a definitive link between AI and the rise in pro se plaintiffs but that the totality of the trends point to a connection. Jeffrey Cohen, an associate professor at Boston College Law School, leads a group of law students assisting pro se litigants in federal court. Since launching the clinic last year with funding from the court system, Cohen said they've noticed the prevalence of AI. Cohen cautions clients to use it with discretion. He sees filings come in way too long, rather than ⁠give the judge a short, plain statement of the claims. "AI can generate so much information in moments, the litigant starts thinking that volume is the key to winning a case," he said. The most troubling piece for him is when would-be plaintiffs ask AI if they have a claim to begin with. If someone asks a chatbot, for instance, whether they have a retaliation claim after being fired, "It's going to say yes," Cohen said. "It never says: No, you don't, you're barred by the statute of limitations." Clients sometimes won't trust his advice if it contradicts what AI told them. U.S. judges have chided ⁠attorneys and pro se litigants almost 1,000 times in recent years for improper AI use, according to a database from Damien Charlotin, a senior research fellow at HEC Paris. In a recent order, U.S. District Judge Marc Treadwell in Georgia ordered a pro se plaintiff to disclose AI use and verify the accuracy of his filings. Treadwell noted, "The Court has seen a marked increase in the use of artificial intelligence ('AI') by unrepresented parties, who may not understand fully their pleading obligations." Judges around the country have issued similar warnings. For Silverberg, her biggest wake-up ⁠call came when a judge called her out for hallucinated case citations in a preliminary injunction motion and threatened sanctions. Now, she reviews everything much more carefully and always asks AI if it produced the best version of a document. Her AI-legal journey - which has also included a state-court eviction case and divorce-related proceedings - has inspired her to help others. She self-published a book on pro se litigation and enrolled in January in Purdue Global Law School, an online program that would enable her to practice law in California. It took a year to properly serve the seven defendants in the case against her landlord. Now it's grinding along. "Of course they're trying to wear me out," Silverberg said. "I've got all the time in the world." Reporting by Sara Randazzo Our Standards: The Thomson Reuters Trust Principles., opens new tab * Suggested Topics: * Government * Judiciary * Legal Innovation * Product Liability Opinions expressed are those of the author. They do not reflect the views of Reuters News, which, under the Trust Principles, is committed to integrity, independence, and freedom from bias. Sara Randazzo Thomson Reuters Sara Randazzo is a columnist for Reuters covering the business and power dynamics driving the nation's legal system. Based in Los Angeles, she previously reported on law firms and litigation for The Wall Street Journal and legal trade publications in California and New York. Contact: [email protected].

People without lawyers are using AI to flood courts with lawsuits

Serving tech enthusiasts for over 25 years. TechSpot means tech analysis and advice you can trust. In a nutshell: It's a case of another day, another problem caused by the proliferation of generative AI. More and more people are using the technology to file lawsuits in which they represent themselves, and, unsurprisingly, courts are being clogged with many of these frivolous suits. The New York Times highlights the case of Donald Sauve, a Minnesota man who sued his ex-wife, her lawyer, and a state judge after an earlier legal challenge was rejected as frivolous. His first handwritten complaint, which sought $275,000 in damages, was dismissed in less than a month for lack of jurisdiction. Three months later, Sauve returned with help from ChatGPT and Claude. This time, his complaint was neatly typed and accompanied by 50 additional filings, including what he called a "case law synthesis" supporting his claim. The case was dismissed, again, but before that happened, every filing had to be read, captioned by a clerk, and entered into the public docket. Sauve's example is a perfect illustration of the problem courts are now facing. AI has made it easier for people without lawyers to produce documents that look and sound like they were prepared by legal experts, even when the underlying claims are weak, confused, or baseless. In Sauve's case, Judge Patrick J. Schiltz, chief of Minnesota's US District Court, ordered that any further filings would be "shredded without any additional notice." Lawsuits filed by people representing themselves rather than using a lawyer are called pro se cases. They are common in federal courts, but most come from prisoners using prison law libraries to challenge conditions, convictions, or alleged civil rights violations. The AI-driven concern focuses more on non-prisoner pro se litigants. These are usually ordinary people who either can't afford legal help or believe they can argue their own case, and can now use chatbots to generate large volumes of court-ready paperwork. A new working paper from MIT's Anand Shah and USC's Joshua Levy examined the scale of the issue. Looking at more than 4.5 million non-prisoner federal civil cases from fiscal 2005 through fiscal 2026, along with 46 million PACER docket entries, the researchers found that non-prisoner pro se cases rose from a long-term average of around 11% to 16.8% in fiscal 2025. The study also found that pro se cases are creating more work once they enter the system. The volume of docket entries per court generated by these cases in the first 180 days had risen 158% above the pre-AI average by 2025. In a sample of 1,600 complaints from 2019 to 2026, more than 18% of 2026 complaints were flagged as likely containing AI-generated text. The counterargument here is that AI could help people who can't afford attorneys access a legal system that is often intimidating and expensive. Judge Michael Y. Scudder of the Seventh US Circuit Court of Appeals wrote this year that the technology offers "great promise" for improving access to justice. But courts are already warning that self-represented litigants remain responsible for AI-generated errors. In January, the Seventh Circuit said accuracy and honesty still matter after a pro se filing appeared to contain AI-generated false citations. As we reported last year, AI-generated legal filings were already making a mess of the judicial system by producing fake cases, phantom precedents, and fabricated citations. Today, the problem has moved beyond lawyers cutting corners to almost anyone with access to a chatbot.

Random People Armed with AI and No Lawyer Are Reportedly Filling Judicial Dockets with Lawsuits

Don’t play innocent. If you’re a non-lawyer in the 2020s, you’ve at least had the passing thought that you could use an LLM to help you generate a killer lawsuit against someone who pissed you off. Or at least now I know it’s not just me. Thanks to AI, plaintiffs representing themselves, also known as “pro se†plaintiffs, are changing the legal landscape for the worse, according to a new study by MIT’s Anand Shah and USC’s Joshua Levy, reported on by the New York Times on Monday. The study has not yet been peer reviewed. It says that since the rollout of widely available LLMs, 18 percent of pro se filings now contain what the authors have deemed AI-generated text. Perhaps consequently, “the total volume of pro se docket entries per court in the first 180 days of a case has grown by 64% on average across the post-AI period,†the study finds. Typically, pro se filings come from prisoners working on their cases from behind bars, but the study notes that “national non-prisoner pro se filing share rose sharply from its approximately 11% historical steady state to 16.8% in fiscal year 2025, a gain that has no precedent in 25 years of administrative records.†According to the Times, pro se plaintiffs lost 96% of their cases from 1998-2017. The Times is largely spotlighting frivolous lawsuits generated with AIâ€"and what a waste of time it is for the courts to painstakingly read and process all these slop-filled filings. A Minnesota federal judge named Patrick J. Schiltz, called it “an existential threat to the federal courts.†To illustrate their point, the Times interviewed a man who uses AI to generate lawsuits. This person gave the paper his name, and allowed himself to be photographed for the story. Courts have alleged some unsavory things about this person, and the Times says he lives in his car. He is, to use one of the president’s favorite terms, straight from central castingâ€"so much so that the Times' story borders on, well, mean. I can’t dispute that AI lawsuits sound like a massive problem. At the same time, lawsuits are often the only weapon downtrodden Americans haveâ€"a substitute for institutions and politicians that actually help make us whole when we're harmed and it's not our fault. Part of me can’t help but long to read a David and Goliath story about a rando armed with Claude who bootstraps their way to some life-changing, ten-figure legal victoryâ€"presumably after using the LLM to figure out how to argue a case in a courtroom as well.

Artificial intelligence floods court dockets with home-brewed lawsuits

The complaint Donald Sauve submitted in Minnesota last year was a familiar type in the nation's federal courts. In legal parlance, Sauve filed "pro se," Latin for "for oneself," meaning he had no lawyer as he sued his ex-wife, her lawyer and a state judge who had rejected one of his earlier legal challenges as "frivolous." In a handwritten scrawl, he previously filed a suit asking for $275,000 in damages, claiming he had been unlawfully deprived of his home. It took less than a month for Judge Jerry W. Blackwell to dismiss the case for lack of jurisdiction. Three months later, Sauve was back. This time he had help. Using ChatGPT and Claude, Sauve filed a new complaint in federal court. This time, it was a neatly typed document accompanied by 50 additional filings including a "case law synthesis" of legal research he said backed up his claim. In an interview, Sauve said artificial intelligence had provided "the only path forward" for his case. "Knowledge is power," he said. Federal judges and legal experts said they are increasingly seeing filings like Sauve's flooding court dockets and clogging an already overburdened system as AI supercharges pro se litigation -- even as it opens up the legal system to people who might not otherwise be able to afford to bring a case. The eventual outcome for Sauve was the same. In September, two months after he filed, Judge Patrick J. Schiltz, chief of Minnesota's U.S. District Court, dismissed his suit again, this time in a 14-page opinion that found he had failed to clearly state a claim. But first, each one of Sauve's filings had to be read, captioned by the clerk and entered on the public docket. Schiltz entered an order that any further filings would be "shredded without any additional notice." "A litigant cannot dump hundreds of pages of documents on a court and expect the court to sift through them to find facts or arguments that might support claims against a defendant," Schiltz wrote, as he dismissed the case. In interviews, judges and experts said that the use of AI by pro se litigants also offers potential upsides. "AI presents great promise for enhancing access to justice for those without the resources to retain counsel or to represent themselves effectively," wrote Judge Michael Y. Scudder of the 7th U.S. Circuit Court of Appeals this year, when ruling on a pro se case. And some federal judges are also discussing ways that AI could be used responsibly to help with workflow in their own chambers, noting that it's possible that AI could someday be used to help clerks to read and assess a larger number of filings. In the meantime, however, many judges emphasized the seriousness of the immediate workload problem created by AI-enabled pro se filings. Schiltz, who declined to discuss any particular case, characterized the overall problem as "an existential threat to the federal courts." The arrival of AI has caused the number, length and complexity of pro se filings to "increase dramatically," he said. "There's just no end in sight, and no satisfactory solution in sight either." 'Something Has to Give' Each year, U.S. District Courts handle roughly 300,000 new lawsuits; another 42,000 new cases are filed in the courts of appeal. One third of that combined caseload comes from pro se litigants, according to data compiled by the Administrative Office of the U.S. Courts, which helps oversee the federal court system. Many of them are prisoners using law libraries to assert their civil rights or otherwise challenge prison conditions. Others are ordinary people who either can't afford a lawyer or believe themselves to be their own best advocate. Between 1998 and 2017, pro se plaintiffs lost 96% of the cases they brought. But judges, lawyers and academics say the volume of filings by pro se litigants has risen dramatically alongside AI's widespread adoption. The proportion of pro se cases filed by nonprisoners increased from 11% of all civil cases five years ago to 16.8% in 2025, according to a new study by two doctoral candidates that has not been peer reviewed. The study found that much of the increase comes from the use of AI by pro se plaintiffs. The number of pro se complaints flagged as likely containing AI-generated text rose from virtually zero in 2019 to more than 18% in 2026, the study found. "Judges still only have 24 hours in a day," said Anand V. Shah of the Massachusetts Institute of Technology and one of the study's authors. "Something has to give at some point." For litigants, the power of generative AI lies in its ability to turn a few short prompts from a user into lengthy documents with headers, citations and other earmarks of a legitimate legal brief. Steven Donohue, a staff attorney for the U.S. District Court for the District of Minnesota in charge of reviewing pro se filings, said he observed a roughly 50% uptick in filings from nonprisoners starting around March 2025. All sorts of pro se cases have been on the rise, Donohue said, including lawsuits alleging false arrest, malicious prosecution and messy domestic disputes involving divorce. In some instances, the cases involved "the bread and butter of state court," now filed in federal court with the help of AI-inspired applications of federal law. "Every eviction action could turn into a Fair Housing Act violation," he said. With careful human oversight, the output of AI can sometimes rival the work of a legal professional, at least for simple matters like drawing up a lease. But in more complex cases, as well as matters that might not be appropriate for a lawsuit, AI's well-established tendency to sometimes flatter and fabricate can cause it to churn out pages of quasi-legal boilerplate that lacks legal merit. Judges "look for truth," said Judge Joshua D. Wolson of the Eastern District of Pennsylvania at a judicial conference in May, speaking to the broader implications of AI. "When the ability to make things that look like truth but aren't -- the cost goes down, the quality goes up -- that's a real challenge for us as courts." Anthropic, the company that owns Claude, did not respond to a request for comment. A spokesperson for OpenAI, the company that owns ChatGPT, declined to comment. Their product's terms of use state that users own and bear the responsibility for its output, which should not be used as "a substitute for professional advice." (The New York Times sued OpenAI and Microsoft in 2023, claiming copyright infringement of news content related to AI systems. The two companies have denied those claims.) Fighting Alone Despite the drawbacks, judges and advocates said AI could be democratizing for the legal system -- opening the courts to people who might otherwise not be able to afford lawyers. "Used appropriately, it could be an incredibly powerful tool for someone who believes themselves to have been wronged and has a good faith belief in entitlement to redress," Donohue said. Decades before AI, pro se filings have led to monumental changes in the law. In 1963, a handwritten petition to the Supreme Court by Clarence Earl Gideon, a 52-year-old convict in a Florida state prison, established a constitutional right to counsel for felony cases in state courts. Sateesh Nori, a legal aid lawyer in New York for 20 years, said he embraced AI two years ago after concluding that legal aid resources in New York were failing to meet the city's needs. Despite a landmark law in 2017 that guaranteed free legal representation to low-income tenants facing eviction, Nori noted city data showing that as many as 50% still go to court without representation. "The real problem is: How come these people don't have another way, other than using AI," he said. Experts said there is little the courts can do to stem pro se filings even if they wanted to. People have a right, after all, to file lawsuits if they believe they have a claim. Peter Kaplan, a spokesperson for the Administrative Office of the U.S. Courts, said his office was "aware of this issue" and "is gathering information" on its effects on the legal system. In recent months, some courts have started issuing standing orders warning prospective pro se filers that using generative AI could expose them to penalties. When judges have run into filings with problematic hallmarks of AI, such as citations to case law that look real but are in fact fictional, many have been lenient to filers who are not lawyers. But some have flashed exasperation, dismissed cases and even issued substantial fines. In March, Judge Virginia Kendall, a federal judge in Illinois, fined a litigant $1,500 after concluding she had twice submitted a "fake case," littered with hallucinated citations in violation of court rules. "This wastes both the parties' and the court's time attempting to locate nonexistent cases and unpack made up factual assertions," she wrote. States have also started to explore legislation that would make AI companies liable if their chatbots were found to have handed out legal advice in place of a lawyer or other licensed professional but such measures have not yet been adopted. For some aggrieved litigants, pushback is unlikely to convince them to give up their AI-empowered legal arsenal. Sauve, the pro se filer from Minnesota, said he continues to pursue efforts to regain his old home. He lost possession of it following a messy divorce, which included a finding from a judge that he abused his ex-wife and one of his children. He denies that allegation. "They call me 'frivolous.' That appears to be a way that the court is protecting itself," he said. "OpenAI told me this, and I think Claude will confirm it as well," he said of his belief that the law is on his side. "I did a lot of research into this whole situation." Sauve is 69 years old and currently living out of his car in Mora, Minnesota. He continues to pursue various forms of legal redress from a supermarket coffee shop. With the help of AI, he plans to soon make more "SCOTUS-grade filings," he said, bringing new lawsuits to the state supreme court, a federal appeals court and a county court as well.

People are using AI to defend themselves in US courts: Why it's a problem

Donald Sauve walked into a federal court in Minnesota with nothing but a handwritten complaint and a grievance against his ex-wife, her lawyer and a state judge. His claim was that he had unlawfully been deprived of his home. This case was dismissed almost immediately for lack of jurisdiction. Still wanting justice to be served, he returned to the court 3 months later but this time he was armed with the powers of the ever so helpful AI Chatbots. Also read: Starlink has a radio wave problem that laser communication can fix, here's how Sauve came back with the an impressively drafted complaint with the help of ChatGPT and Claude, 50 additional filings, and a "case law synthesis" based on the legal research he conducted that supposedly supported his claims. The documents were professionally done. They seemed credible. And yet, they did not make a difference at all and were tossed away once more after being the basis for a judicial ruling consisting of 14 pages. A handwritten complaint was rejected after some weeks. An AI version took much more work from the court before coming to the same decision. Every filing had to be read, logged, captioned by a clerk, and entered into the public record before Judge Patrick Schiltz could finally rule that Sauve had failed to clearly state a single claim. The point is worth mentioning here. Sauve made 50 filings and a case law synthesis based on legal research. And there were no legitimate claims to substantiate a case. In doing so, it reveals the fundamental illusion that is being perpetrated. While AI enabled Sauve to structure his claim as a legally legitimate document, complete with proper citations, formatting, and legal jargon, what it could not offer him was the substance of the claim itself. Far from being an openly laughable submission, what the software created was something far worse, a veneer of legal legitimacy that belied a total lack of substance behind it. Also read: More code, more vulnerabilities, more jobs: How AI is reshaping cybersecurity hiring And such a disjuncture between legal substance and superficial legitimacy lies at the center of a crisis brewing in America's federal courts. While federal courts have long been hospitable to individuals filing "pro se," meaning without legal representation, there is a new reality to consider. What once required months of self-education in a law library can now be generated in minutes. The volume, and the problem, has grown accordingly. Federal district courts are faced with about 300,000 civil cases annually, while another 42,000 civil lawsuits are filed before appellate courts. As many as one-third of these cases have been initiated by pro se litigants. According to the judges and attorneys, this number is on the rise - not because there are more complaints that require judicial attention, but rather because of how easily AI can generate legal-looking documents based on questionable premises. The consequences are not abstract. Every filing, however meritless, demands judicial time and staff resources. Judge Schiltz eventually ordered that any further filings from Sauve would be destroyed without notice. A court, he wrote, cannot be expected to excavate hundreds of pages of documents searching for facts that might support a claim. However, the use of AI is neither purely evil nor illegal. On the contrary, both legal professionals and judges admit that AI has much potential for democratizing law and making it accessible for everyone who could otherwise not afford to hire a lawyer. The problem is that the entire structure of the justice system was not designed to accommodate such changes.