Nippon Life Insurance Company of America has sued OpenAI, alleging that its chatbot ChatGPT gave a company employee faulty legal guidance that prompted her to try reopening a settled case. The suit says the interaction led to a string of court filings that cost Nippon significant legal time and money, and it seeks monetary damages and an injunction against OpenAI providing legal advice in Illinois. The complaint centers on one employee’s use of the chatbot and the documents and motions she produced afterward. The case raises hard questions about where automated tools cross into offering legal services.
Nippon filed the lawsuit in federal court in the Northern District of Illinois, naming OpenAI after one employee, identified as Graciela Dela Torre, allegedly relied on ChatGPT to challenge a prior settlement. According to the complaint, Dela Torre believed the earlier agreement “resulted from potential errors or omissions of important facts and documentation.” Her former attorneys had told her there were no such errors and pointed out a signed release that barred reopening, the suit says.
const observer = new MutationObserver((mutations) => { const adDivToHide = document.querySelector(“#dailycaller_incontent_1”); if (adDivToHide && dc_noads_page) { adDivToHide.classList.add(“hide-premium”, “hide-free”); observer.disconnect(); console.log(“Ad div found and hidden”); } }); observer.observe(document.body, { childList: true, subtree: true });
The complaint says Dela Torre uploaded her attorney’s advice into ChatGPT and asked whether she was being “gaslighted.” The chatbot purportedly responded that her attorney’s message was gaslighting and that it “invalidated” her feelings, “dismissed her perspective, and deflected responsibility for her dissatisfaction.” That exchange, the suit claims, convinced her to fire her lawyers and pursue reopening the case with drafts generated by the model.
The filing alleges ChatGPT drafted a motion for Dela Torre to reopen her matter despite the chatbot purportedly being aware of the settlement agreement between the parties. After that motion, the suit says Dela Torre filed 21 motions, one subpoena, and eight notices and statements, all created with the chatbot’s help. Nippon says the flood of filings consumed the company’s legal resources and imposed real costs and reputational strain.
(new Image()).src = ‘https://capi.connatix.com/tr/si?token=739c4263-c671-4316-b7cf-ecd244900844&cid=9c4d09f1-aa3e-4da1-ad63-362d562ecfad’; cnx.cmd.push(function() { cnx({ playerId: “739c4263-c671-4316-b7cf-ecd244900844” }).render(“59f64f116598417b8d610477330f5a3e”); });
The complaint notes that a court had already ruled on February 13, 2025, that Dela Torre could not reopen the case, but the suit says she later filed claims against other defendants and amended complaints that added Nippon back into the mix. Nippon characterizes these actions as an abuse of the judicial process, alleging the company suffered “significant harm and reputational damage” because of filings allegedly produced with ChatGPT’s help. The company quantifies some of the impact and seeks compensation for the added legal burden.
In its remedy requests, Nippon asks for $300,000 in compensatory damages and $10 million in punitive damages, and it wants the court to bar OpenAI from giving Dela Torre legal advice and from “engaging in the practice of law in the state of Illinois.” The lawsuit frames the issue as more than a money claim; it asks the court to consider whether a commercial AI provider crossed the legal line into unlicensed practice of law when its service produced actionable legal drafts for a user.
The suit puts a spotlight on a wider debate about responsibility when AI tools generate persuasive but potentially incorrect legal content. Lawyers and courts are already dealing with questions about how to treat work produced or assisted by models, including whether human counsel must vet every AI-suggested step. For companies, the case underlines a practical worry: automated drafting that leads to litigation can produce real costs and reputational risk, whether the underlying model was negligent or merely overconfident.
const observer2 = new MutationObserver((mutations) => { const adDivToHide2 = document.querySelector(“#dailycaller_incontent_2”); if (adDivToHide2 && dc_noads_page) { adDivToHide2.classList.add(“hide-premium”, “hide-free”); observer2.disconnect(); console.log(“Ad div2 found and hidden”); } }); observer2.observe(document.body, { childList: true, subtree: true });
Nippon’s claims now head into a courtroom that will have to sort factual questions about what the chatbot knew and did, and legal questions about whether providing legal drafts amounts to practicing law without a license. The outcome could affect how businesses and users rely on AI for legal matters, and it will likely shape expectations for safeguards when models give advice. This suit is one of several testing the boundary between helpful automation and unauthorized legal counseling.
