Sam Altman
TechCrunch/Flickr

An Illinois disability dispute that had already been settled is now at the centre of a new legal fight, after an insurer alleged a ChatGPT chatbot effectively stepped into a lawyer's role and fuelled fresh litigation. The lawsuit claims the bot persuaded Graciela Dela Torre, from Des Plaines, to part ways with her actual attorney and press forward with filings that relied on invented case law.

Court papers say the saga began with a workplace injury claim, then escalated into a wave of motions and a second lawsuit that the insurer argues should never have existed. The complaint targets OpenAI, the company behind the chatbot, and seeks to recoup mounting defence costs as well as punitive damages.

How ChatGPT Allegedly Replaced Her Lawyer In An Illinois Disability Dispute

Dela Torre and her lawyer initially pursued a disability claim against Nippon Life Insurance Company after she suffered carpal tunnel and tennis elbow on the job in August 2019. She later stopped qualifying as disabled in November 2021, then sued and ultimately obtained a settlement, the court papers say.

As part of that agreement, she waived any future claims against the insurer in January 2024. When she tried to revisit the matter about a year later, her lawyer allegedly told her it could not be reopened, prompting her to seek guidance from ChatGPT and ask whether she had been 'gaslighted' by her attorney.

The insurer's lawsuit stresses that the chatbot is not licensed to practise law, even while pointing to its performance on legal testing. It states: 'ChatGPT is not an attorney. Although it was able to pass the Uniform Bar Examination with a combined score of 297, it has not been admitted to practice law in the State of Illinois or in any other jurisdiction within the United States.'

Fabricated Case Law And Dozens Of Court Filings Flood The Docket

After consulting the chatbot, Dela Torre submitted a pro se filing on 22 January 2025 seeking to reopen the settled case, the insurer alleges. A judge ruled on 13 February 2025 that she could not reopen the case, but the filings did not stop there, according to the complaint.

Instead, the court papers say she went on to bring a new suit against Nippon that remains pending. The insurer alleges ChatGPT produced at least 44 filings connected to her efforts, including a document citing the fabricated case 'Carr v. Gateway, Inc. 9'. The complaint argues that the supposed precedent is nowhere to be found outside the bot's output and Dela Torre's submissions. It states: 'It only exists in Dela Torre's papers and the "mind" of ChatGPT.'

The lawsuit also itemises the volume of documents submitted to the court after the failed attempt to reopen the settled case. It states: 'Following the motion to reopen the lawsuit, Dela Torre filed 21 motions, one subpoena, and eight notices and statements onto the docket. All of Dela Torre's motions, subpoenas and notices were compiled and drafted by ChatGPT.'

@masstortmichael

A new lawsuit was filed in Chicago this week that I’ve been thinking about since I read the complaint. 𝐍𝐢𝐩𝐩𝐨𝐧 𝐋𝐢𝐟𝐞 𝐈𝐧𝐬𝐮𝐫𝐚𝐧𝐜𝐞 𝐂𝐨𝐦𝐩𝐚𝐧𝐲 𝐨𝐟 𝐀𝐦𝐞𝐫𝐢𝐜𝐚 𝐬𝐮𝐞𝐝 𝐎𝐩𝐞𝐧𝐀𝐈 in the Northern District of Illinois, Case No. 1:26-cv-02448. The core allegation: ChatGPT practiced law without a license. Here’s the short version of what the complaint says happened. Nippon settled a long-term disability lawsuit in January 2024. The claimant signed a full release. Case dismissed with prejudice. A year later, she changed her mind. Her attorney told her the release was enforceable and the case was closed. She uploaded his response to ChatGPT and asked if she was being gaslighted. ChatGPT said yes. Then ChatGPT drafted her motions, generated her legal arguments, conducted her legal research, and helped her file them in federal court. One filing cited a case that doesn’t exist. It only appears in ChatGPT’s output and her court papers. By the time Nippon filed this new lawsuit, she had submitted more than 60 documents across two cases, nearly all drafted with ChatGPT’s assistance. Nippon says the cost of defending a settled case has reached approximately $300,000. The three claims are tortious interference with a contract, abuse of process, and unlicensed practice of law under Illinois statute. Nippon is seeking $10 million in punitive damages and a permanent injunction barring OpenAI from practicing law in the state. I’m not here to say who’s right. This case is three days old and OpenAI hasn’t responded. But the legal question sitting at the center of this complaint isn’t a hypothetical anymore. When an AI drafts your pleadings, analyzes your case, tells you to fire your lawyer, and generates the legal strategy you take to federal court, is that the practice of law? And if it is, does the company that built the tool bear responsibility for what it produces? Courts are going to have to answer that. This may be one of the first cases where they try. I’ll be watching it closely. Full complaint is publicly available on PACER if you want to read it yourself. #masstortmichael #openailawsuit LexHive DemandLane #NipponLawsuit Not Legal or Medical Advice

♬ original sound - masstortmichael

Why Nippon Is Seeking Over £7.7m From OpenAI

Nippon alleges it has spent $300,000 (approximately £224,000) defending against the revived litigation and related filings, and is seeking that sum plus $10 million (approximately £7.5 million) in punitive damages.

The complaint further characterises Dela Torre's motives in blunt terms, alleging a pattern that goes beyond legitimate legal aims. It states: 'Dela Torre's conduct throughout both lawsuits reveals a state of mind driven by sustained animosity rather than any objective legal purpose. She has consistently portrayed herself as impoverished, accused Nippon of financially starving her, and repeatedly blaming Nippon for her financial hardship.'

Dela Torre is described in the reporting as a senior logistics coordinator, and she is not named as a defendant in the insurer's lawsuit against OpenAI. OpenAI has rejected the allegations, with a spokesperson quoted as saying: 'This complaint lacks any merit whatsoever.'

More details are set out in the insurer's claims and supporting court papers, including the request for punitive damages tied to the alleged misuse of AI-generated legal work.

Writer's pick