{"meta":{"exportedAt":"2026-10-09T06:45:28.407Z","formatVersion":2,"selection":{"q":"professional_harm","system":"","harm":"","context":"","country":"","role":"","relation":"core","evidence":"","year":"2025","response":"","severity":"","verification":"","view":"incidents","sort":"added"},"totalIncidents":8,"coverage":{"cases":8,"countries":3,"languages":2,"unknownLocation":2,"locationPending":0,"unknownLanguage":0,"unknownDate":0,"lawsuits":0,"regulatory":3,"minors":0,"coreRelations":8,"contextualRelations":1,"mixedRelations":1,"unknownRelations":0,"relationPending":0,"relationUnknown":0},"countingNote":"Distinct public cases in this selection. People counts apply within individual cases only; cross-case person overlap has not been resolved. No population incidence estimate.","affectedCountNote":"Interpret person counts with affectedCountStatus and the reported effects. Unquantified zeros are placeholders, not a measured zero.","source":"AI incidents","publisher":"NOPE","url":"https://nope.net/incidents","license":"CC BY 4.0"},"incidents":[{"id":"2025-uk-upper-tribunal-asylum-appeal-chatgpt-grounds-fake-cases-invented-quotes-sra-self-report","caseFacts":{"claims":[{"id":"c1","status":"documented","evidence":[{"locator":"the grounds of appeal were completely infected by the use of AI. In addition to the citation of fake cases, and paragraphs from cases which are not contained in those judgments, the Grounds of Appeal also quote paragraphs from the decision of the First-tier Tribunal which do not appear in that decision.","relation":"supports","source_id":"s1"},{"locator":"This was pure invention, which was not to be found anywhere in Judge Balroop’s decision, yet criticism of it formed the basis for one aspect of ground (3) in the grounds of appeal.","relation":"supports","source_id":"s1"},{"locator":"In this case, Mr Badar identified at least 18 separate errors in the grounds of appeal and withdrew three of the four grounds.","relation":"supports","source_id":"s1"}],"assertion":"The appellant's then solicitors drafted grounds of appeal that her counsel accepted were completely infected by the use of AI, citing fake cases and paragraphs not contained in the cited judgments and quoting passages from the First-tier Tribunal decision that do not appear in it; the tribunal found one purported quotation to be pure invention.","causal_attribution":"Established by the tribunal's own findings and its record of counsel's concession in the published decision. The Gazette repeats the decision and adds no independent origin."},{"id":"c2","status":"documented","evidence":[{"locator":"Both were produced by a junior member of staff, namely a paralegal, using ChatGPT, which itself raises significant concerns about client confidentiality and legal privilege.","relation":"supports","source_id":"s1"},{"locator":"This occurred because Sarker Solicitors did not check the accuracy of the grounds of appeal, did not consider the grant of permission and did not check the accuracy of the skeleton argument.","relation":"supports","source_id":"s1"},{"locator":"They accepted that AI was used to generate the grounds of appeal filed in the FtT and the skeleton argument filed in the UT.","relation":"supports","source_id":"s1"},{"locator":"The grounds of appeal and skeleton argument were drafted by an unnamed paralegal working for east London firm Sarker Solicitors.","relation":"supports","source_id":"s2"}],"assertion":"The tribunal found that both the grounds of appeal and the skeleton argument were produced by a junior member of staff, a paralegal, using ChatGPT, and that the firm did not check their accuracy; the solicitors had accepted that AI was used to generate both documents.","causal_attribution":"The product identification and the failure to check are the tribunal's findings in the decision, recorded with the solicitors' acceptance that AI generated the documents. The Gazette's statement that the drafter was an unnamed paralegal relays the decision (same independence group)."},{"id":"c3","status":"documented","evidence":[{"locator":"A hearing was listed before Deputy Upper Tribunal Judge Hobbs on 17 September 2025 to consider whether Judge Balroop had erred in law.","relation":"supports","source_id":"s1"},{"locator":"Judge Hobbs adjourned the hearing, with directions which included the following:","relation":"supports","source_id":"s1"},{"locator":"We note that issues (2) to (4), which occupied a great deal of time at the hearing before us, only arose because of Sarker Solicitors’ inappropriate use of AI.","relation":"supports","source_id":"s1"},{"locator":"The appellant withdrew the original grounds (1), (2) and (4).","relation":"supports","source_id":"s1"},{"locator":"However, ground (3) discloses no error of law.","relation":"supports","source_id":"s1"},{"locator":"The appellant’s appeal therefore fails.","relation":"supports","source_id":"s1"},{"locator":"The judge declined to set aside permission to appeal but dismissed the appeal in any case. The Home Office was invited to file and serve any wasted costs application on Sarker Solicitors within 28 days.","relation":"supports","source_id":"s2"}],"assertion":"The error-of-law hearing listed for 17 September 2025 was adjourned, three of the five issues at the final hearing, which occupied a great deal of its time, arose only because of the solicitors' use of AI, the appellant withdrew three of her four original grounds, and her appeal failed after her remaining and amended grounds were found to disclose no error of law.","causal_attribution":"Procedural history and disposition are directly established by the decision. The decision attributes the adjournment and the wasted hearing time to the solicitors' misuse of AI, and states separately that the surviving ground failed on its merits, so the dismissal itself is not attributed to the AI output."},{"id":"c4","status":"documented","evidence":[{"locator":"Two individuals have reported themselves to the Solicitors Regulation Authority and so we will say nothing about the professional conduct issues involved.","relation":"supports","source_id":"s1"},{"locator":"The respondent shall file and serve on Sarker Solicitors any wasted costs application, or confirmation that costs are not being pursued, within 28 days of this decision being sent to the parties.","relation":"supports","source_id":"s1"},{"locator":"Two individuals involved in filing court documents using ChatGPT which were riddled with mistakes and invented quotes have reported themselves to the Solicitors Regulation Authority.","relation":"supports","source_id":"s2"}],"assertion":"The decision states that two individuals have reported themselves to the Solicitors Regulation Authority, and directs the Home Office to file any wasted costs application against the firm within 28 days; no costs decision has been taken.","causal_attribution":"The self-reports and the costs directions are stated in the decision itself. The Gazette repeats them from the decision (same independence group), so the claim is documented, not corroborated. Whether the two individuals are the paralegal and the supervising solicitor is not stated."},{"id":"c5","status":"documented","evidence":[{"locator":"It would be misleading to describe them as having been caused by AI, rather than by the solicitors’ misuse of AI.","relation":"supports","source_id":"s1"},{"locator":"This was one of the more extreme examples of the misuse of AI, rather than a mere mis-citation of particular cases.","relation":"supports","source_id":"s1"}],"assertion":"The tribunal said it would be misleading to describe the irregularities as caused by AI rather than by the solicitors' misuse of AI, while describing the case as one of the more extreme examples of the misuse of AI.","causal_attribution":"The tribunal's own qualification, preserved as contrary evidence on causation: the decision places responsibility with the solicitors, and the Home Office's characterisation (at best negligent, at worst reckless) is relayed in the decision as a submission."}],"effects":[{"label":"Appellant's protection appeal pursued on AI-generated grounds containing fake cases and invented quotations; hearing adjourned and three of four grounds withdrawn","claim_id":"c3","direction":"negative"},{"label":"Two individuals self-reported to the Solicitors Regulation Authority; wasted costs application against the firm invited","claim_id":"c4","direction":"negative"}],"sources":[{"id":"s1","url":"https://tribunalsdecisions.service.gov.uk/utiac/ui-2025-002601","kind":"court_record","access":"read","language":"en","translation_note":"","independence_group":"ut-decision-ui-2025-002601"},{"id":"s2","url":"https://www.lawgazette.co.uk/news/duo-self-report-to-sra-over-grounds-of-appeal-infected-with-ai/5128148.article","kind":"news_report","access":"read","language":"en","translation_note":"","independence_group":"ut-decision-ui-2025-002601"}],"version":1,"ai_roles":["others_use"],"contexts":["justice"],"unknowns":["The date on which the paralegal generated the grounds of appeal is not stated; it falls between the First-tier decision of 23 April 2025 and the hearing listed for 17 September 2025.","Whether the two individuals who self-reported to the Solicitors Regulation Authority are the paralegal and the supervising solicitor is not stated, and the outcome of any SRA consideration is unknown.","Whether the Home Office pursued wasted costs, and any amount, was not decided in the published decision.","What prompts were given to ChatGPT, and whether client material was entered into it, is not described; the tribunal only notes that the use raises confidentiality and privilege concerns.","Whether the appellant's appeal would have succeeded on properly drafted grounds cannot be known; the tribunal found the surviving and amended grounds disclosed no error of law."],"geography":{"basis":"The decision is an Upper Tribunal (Immigration and Asylum Chamber) decision heard at Field House, London, in an appeal against a UK Home Office refusal; the solicitors are described by the Gazette as an east London firm. The appellant is a Namibian national whose appeal concerns her removal from the UK, so her country is recorded as GB on the basis of the proceedings in which she was harmed, and Namibia is not recorded because no harm occurred there.","court_countries":["GB"],"event_countries":["GB"],"affected_person_countries":["GB"]},"publication":{"basis":"The published Upper Tribunal decision (tribunalsdecisions.service.gov.uk, UI-2025-002601) and the Law Society Gazette report of 8 October 2026 were read in full on 2026-10-09. The decision records what ChatGPT generated (fake cases, misattributed paragraphs, invented quotations from the First-tier judge) and ties that output to the adjourned hearing, the wasted hearing time, the withdrawn grounds, the regulator self-reports and the costs directions. The appellant is anonymised by tribunal order and is described only by nationality and appeal type; the firm is named because the operative costs directions are addressed to it; the paralegal is unnamed. The tribunal's statement that the solicitors' misuse, not AI, caused the irregularities is preserved.","reviewed_on":"2026-10-09"},"ai_involvement":{"basis":"The Upper Tribunal's published decision finds that the grounds of appeal and the skeleton argument filed on the appellant's behalf were produced by a paralegal using ChatGPT, and describes what that output contained: citations of fake cases, paragraphs not found in the cited judgments, and purported quotations from the First-tier judge's decision that the tribunal calls pure invention. The decision connects that output to the harm: the September 2025 hearing was adjourned after the Home Office queried the citation errors, three of the five issues at the final hearing, which occupied a great deal of its time, arose only because of the solicitors' use of AI, the appellant withdrew three of four grounds, and the tribunal invited a wasted costs application and noted two self-reports to the regulator. The FtT judge granted permission on those grounds while noting the misquotations, and the Upper Tribunal's decisions not to set aside permission and to refuse the new grounds were taken in the express context of the misrepresentations. The tribunal itself states that the irregularities were caused by the solicitors' misuse of AI rather than by AI, and the decision records the solicitors' own acceptance that AI was used. The firm's acceptance and the tribunal's findings come from one record, so the status is supported rather than corroborated by an independent source.","status":"supported"},"person_relations":["acted_on_behalf","informed_decision_about"]},"name":"UK Upper Tribunal: solicitors' ChatGPT-drafted grounds in an asylum appeal cited fake cases and invented quotes; two individuals self-report to the SRA","summary":"A published Upper Tribunal (Immigration and Asylum Chamber) decision issued on 21 September 2026 records that the grounds of appeal and skeleton argument in a Namibian national's protection appeal were produced by a paralegal at her then solicitors using ChatGPT. The tribunal recorded her counsel's acceptance that the grounds cited fake cases, quoted passages absent from those judgments and attributed to the First-tier judge sentences he never wrote, and itself called one purported quotation pure invention. The September 2025 error-of-law hearing was adjourned, the appellant later withdrew three of her four original grounds, her remaining and amended grounds were found to disclose no error of law, and the tribunal directed the Home Office to file any wasted costs application against the firm. The decision states that two individuals have reported themselves to the Solicitors Regulation Authority; the Law Society Gazette reported the decision on 8 October 2026.","incidentDate":"2025-09-17","incidentEndDate":"2026-09-21","incidentKind":"bounded_series","incidentDatePrecision":"range","exposurePattern":"unknown","reportedDate":"2026-10-06","aiSystem":"ChatGPT, used by a paralegal to generate grounds of appeal and a skeleton argument (as found in the Upper Tribunal decision)","aiProduct":"ChatGPT","aiCompany":"OpenAI","severity":"low","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["legal_harm","professional_harm"],"harmOutcomeSummary":"The tribunal records that the appellant's error-of-law hearing in September 2025 was adjourned, that three of the five issues at the eventual hearing, which occupied a great deal of its time, arose only because of her solicitors' inappropriate use of AI, that she withdrew three of her four original grounds, and that her appeal fails. The decision states that two individuals have reported themselves to the Solicitors Regulation Authority and directs the Home Office to file any wasted costs application against the firm; no costs decision has been taken.","frameworkFacets":[],"causationStatus":"established","participantUsersAffectedMin":0,"otherPeopleHarmedMin":1,"affectedCountStatus":"documented_minimum","affectedCountEvidence":"One person counted: the appellant, whose appeal proceeded on AI-generated grounds she did not draft (the decision describes her as the appellant throughout). The paralegal who used ChatGPT and the supervising solicitor face professional consequences of their own conduct (self-report to the SRA, possible wasted costs) and are not counted as harmed persons; the two self-reporting individuals are not named as the paralegal and the solicitor in the decision.","victimAgeRange":"adult","jurisdiction":"GB","platformType":"chatbot","primarySourceUrl":"https://tribunalsdecisions.service.gov.uk/utiac/ui-2025-002601","primarySourceLabel":"Upper Tribunal (Immigration and Asylum Chamber), UI-2025-002601, decision issued 21 September 2026, published 6 October 2026","firstPublishedAt":"2026-10-09T03:28:56.311373+00:00","updatedAt":"2026-10-09T03:28:56.311373+00:00","scopeVersion":"facts-v3","tags":[]},{"id":"2025-north-carolina-federal-prosecutor-ai-drafted-brief-fabricated-quotations-reprimand","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"This filing misquoted a Fourth Circuit case, falsely attributed quotes to other cases and incorrectly described the court’s decision in several cases that he relied upon to support his argument, Numbers said.","relation":"supports","source_id":"s3"},{"locator":"Numbers' list of problematic conduct includes fabricated quotes, misrepresentations of previous cases","relation":"supports","source_id":"s1"},{"locator":"filed a response with the court including “fabricated quotations and misstatements of case holdings”","relation":"supports","source_id":"s4"}],"assertion":"A response brief filed in December 2025 by Assistant U.S. Attorney Rudy Renfer for the Department of Defense in Fivehouse v. U.S. Department of Defense misquoted a Fourth Circuit case, attributed quotes to cases that did not contain them and misdescribed the holdings of several cases, according to Magistrate Judge Robert T. Numbers II's orders as reported by the News & Observer and Courthouse News.","causal_attribution":"The court attributed the false quotations to the brief Renfer signed; the fabricated authority is the AI-related artifact in this case."},{"id":"c2","status":"reported","evidence":[{"locator":"flagged fabricated quotes and inaccurate descriptions of court cases in a brief Renfer filed earlier that month.","relation":"supports","source_id":"s1"},{"locator":"In a January court filing, Renfer attributed the errors to an \"inadvertent filing of an unfinalized draft.\"","relation":"supports","source_id":"s1"},{"locator":"Fivehouse pointed out the errors in his reply.","relation":"supports","source_id":"s3"}],"assertion":"The self-represented plaintiff, a retired Air Force colonel, identified the fabricated quotes and inaccurate case descriptions in his reply at the end of December 2025, and Renfer then attributed the errors to the inadvertent filing of an unfinalized draft.","causal_attribution":"none asserted"},{"id":"c3","status":"documented","evidence":[{"locator":"the court orders Assistant United States Attorney Rudy E. Renfer to appear and show cause why he should not be sanctioned. The show cause hearing will be held on Tuesday, March 10, 2026","relation":"supports","source_id":"s5"},{"locator":"must appear at a show-cause hearing this week","relation":"supports","source_id":"s4"}],"assertion":"On 2 March 2026 the court ordered Renfer to appear on 10 March 2026 and show cause why he should not be sanctioned under Rule 11(b)(2) and the court's inherent authority, and asked leadership of the U.S. Attorney's Office to attend.","causal_attribution":"none asserted"},{"id":"c4","status":"reported","evidence":[{"locator":"Renfer then admitted to using AI to draft the brief","relation":"supports","source_id":"s1"},{"locator":"he thought he had edited and fact-checked the brief that AI helped draft. In reality, he sent an unfinished draft to his assistant, directing her to file it","relation":"supports","source_id":"s1"},{"locator":"He told the court that he accidentally saved a new file over the original draft of the brief, “panicked,” and used artificial intelligence to catch up.","relation":"supports","source_id":"s3"}],"assertion":"At the 10 March 2026 hearing Renfer said under oath that, after accidentally saving over his draft of the brief, he panicked and used artificial intelligence to draft it, believed he had edited and fact-checked the AI-assisted draft, and in fact sent an unfinished draft to be filed.","causal_attribution":"Renfer connected the fabricated authority to his use of AI; the court later wrote that an attorney who outsources research and writing to generative AI assumes the risk of fictitious authority."},{"id":"c5","status":"reported","evidence":[{"locator":"Renfer lost his job after resigning from the position he had held for 17 years, he said.","relation":"supports","source_id":"s1"},{"locator":"Renfer also told Numbers, who presided over that hearing, that he had submitted his resignation.","relation":"supports","source_id":"s3"},{"locator":"Renfer was reportedly terminated the next day.","relation":"supports","source_id":"s3"},{"locator":"has asked the Department of Justice's Office of Professional Responsibility to investigate","relation":"supports","source_id":"s1"}],"assertion":"Renfer left the U.S. Attorney's Office, where he had worked for 17 years: he told the court on 10 March 2026 that he had submitted his resignation, and Courthouse News reported that he was terminated the next day; the U.S. Attorney referred him to the Department of Justice's Office of Professional Responsibility.","causal_attribution":"The job loss followed the hearing on the AI-drafted brief; the sources do not state the employer's formal grounds."},{"id":"c6","status":"reported","evidence":[{"locator":"On Tuesday, U.S. Magistrate Judge Robert Numbers II signed an order reprimanding Rudy Renfer","relation":"supports","source_id":"s2"},{"locator":"Numbers limited the punishment to a public reprimand.","relation":"supports","source_id":"s2"},{"locator":"Numbers said Renfer “intentionally submitted a brief containing false materials to the court.”","relation":"supports","source_id":"s3"},{"locator":"His loss of employment imposes a financial burden well beyond the types of fines courts typically","relation":"supports","source_id":"s3"}],"assertion":"On 28 April 2026 Magistrate Judge Numbers signed an order publicly reprimanding Renfer, writing that he intentionally submitted a brief containing false materials, that his professional reputation was in tatters, and that his loss of employment was a financial burden beyond the fines courts typically impose, so no fine was imposed.","causal_attribution":"The court tied the reprimand to the submission of fabricated authority produced with generative AI."}],"effects":[{"label":"a federal prosecutor lost his 17-year position and received a public judicial reprimand after filing an AI-drafted brief with fabricated quotations","claim_id":"c6","direction":"negative"},{"label":"a self-represented litigant had to identify fabricated quotations and misdescribed cases in the government's brief against him","claim_id":"c2","direction":"negative"}],"sources":[{"id":"s1","url":"https://www.yahoo.com/news/articles/judge-weighs-next-steps-us-121054156.html","kind":"news_report","access":"read","language":"en","translation_note":"Read in English by the research agent (an AI) on 2026-10-09 from the saved body; no translation was involved and no human reviewer read the text. The News & Observer report of 11 March 2026 on the 10 March show-cause hearing, read in its Yahoo News syndication; HTTP 200.","independence_group":"news-and-observer"},{"id":"s2","url":"https://www.yahoo.com/news/articles/federal-judge-punishes-ex-us-173015519.html","kind":"news_report","access":"read","language":"en","translation_note":"Read in English by the research agent (an AI) on 2026-10-09 from the saved body; no translation was involved and no human reviewer read the text. The News & Observer report of 28 April 2026 on the reprimand order, read in its Yahoo News syndication; HTTP 200.","independence_group":"news-and-observer"},{"id":"s3","url":"https://www.courthousenews.com/judge-rebukes-ex-doj-attorney-for-ai-generated-brief/","kind":"news_report","access":"read","language":"en","translation_note":"Read in English by the research agent (an AI) on 2026-10-09 from the saved body; no translation was involved and no human reviewer read the text. Courthouse News Service report of 28 April 2026; the publisher host returned HTTP 403 to the fetcher, so the body was read from an Internet Archive capture of the same URL.","independence_group":"courthouse-news"},{"id":"s4","url":"https://www.abajournal.com/news/article/doj-leaders-must-explain-why-attorney-used-fabricated-quotes-in-court-filing","kind":"news_report","access":"read","language":"en","translation_note":"Read in English by the research agent (an AI) on 2026-10-09 from the saved body; no translation was involved and no human reviewer read the text. ABA Journal, 9 March 2026, relaying Bloomberg Law reporting of the 2 March 2026 show-cause order; it does not mention AI; HTTP 200.","independence_group":"bloomberg-law"},{"id":"s5","url":"https://www.courtlistener.com/docket/71231282/fivehouse-v-us-department-of-defense/?page=2","kind":"court_record","access":"read","language":"en","translation_note":"Read in English by the research agent (an AI) on 2026-10-09 from the saved body; no translation was involved and no human reviewer read the text. CourtListener RECAP docket for Fivehouse v. U.S. Department of Defense, No. 2:25-cv-00041 (E.D.N.C.), page 2, entries of February to March 2026; HTTP 200. The 28 April 2026 order itself was not available as a free document.","independence_group":"court-docket"}],"version":1,"ai_roles":["own_use"],"contexts":["work","justice"],"unknowns":["Which AI tool Renfer used, and whether the whole brief or only parts of it were generated; the inspected accounts say only that he used artificial intelligence to draft the brief after losing his own draft.","Whether Renfer resigned or was terminated: he told the court he had resigned, and Courthouse News reported that he was terminated the next day.","The outcome of the Office of Professional Responsibility referral, and whether the court found similar fabricated quotations in other filings he signed (the News & Observer mentions errors in some of his other briefs without detail).","Any consequence for the self-represented plaintiff beyond the work of identifying the errors; the inspected sources describe none."],"geography":{"basis":"The brief was filed in the U.S. District Court for the Eastern District of North Carolina and the hearing took place at the federal courthouse in Raleigh (docket entry of 2 March 2026; News & Observer). Renfer worked in that district's U.S. Attorney's Office. No other country is involved.","court_countries":["US"],"event_countries":["US"],"affected_person_countries":["US"]},"publication":{"basis":"Published as a core case on two independent reporting chains that both quote the court's orders and both report the attorney's statements at the 10 March 2026 hearing: a federal prosecutor said under oath that he used artificial intelligence to draft a brief that the court found contained fabricated quotations and misdescribed holdings, and the court responded with a show-cause order (documented on the docket), after which he lost his position and received a public reprimand on 28 April 2026. The affected person is a named public actor in a public court record. The AI tool is not identified and the resignation-or-termination question is recorded as unknown.","reviewed_on":"2026-10-09"},"ai_involvement":{"basis":"Renfer stated under oath at the 10 March 2026 show-cause hearing that he used artificial intelligence to draft the response brief after overwriting his own draft, and that he had not verified the cases it cited (News & Observer, 11 March; Courthouse News, 28 April). The court's 28 April order, as quoted by Courthouse News, treated the fabricated quotations and misdescribed holdings in that brief as the output of generative AI that Renfer failed to check ('An attorney who outsources core research and writing obligations to generative AI assumes the risk that its output will contain fictitious authority'); the News & Observer quotes the same order's statement that 'the submission of fabricated authority will not be tolerated in this Court'. The fabricated authority in the filed brief led to the show-cause order, the hearing, his departure from the office and the public reprimand. The tool is not identified in any inspected source.","status":"supported"},"person_relations":["communicated_with"]},"name":"Former federal prosecutor in North Carolina publicly reprimanded after admitting that a brief with fabricated quotations was drafted with AI; he lost his post","summary":"In December 2025 an Assistant U.S. Attorney in the Eastern District of North Carolina filed a response brief for the Department of Defense in a self-represented veteran's lawsuit over health-insurance coverage. The plaintiff identified fabricated quotations and misdescribed case holdings in it. The attorney first attributed the errors to an unfinalized draft filed by mistake. On 2 March 2026 Magistrate Judge Robert T. Numbers II ordered him to show cause why he should not be sanctioned. At the 10 March hearing the attorney, Rudy Renfer, said under oath that he had accidentally overwritten his draft, panicked and used artificial intelligence to draft the brief, and had not verified the cases it cited; he said he had resigned after 17 years in the office, and Courthouse News reported he was terminated the next day. On 28 April 2026 the court issued a public reprimand instead of a fine, writing that his reputation was in tatters and that his job loss already exceeded the usual fines.","incidentDate":"2025-12-23","incidentEndDate":"2026-04-28","incidentKind":"single_event","incidentDatePrecision":"day","exposurePattern":"single_interaction","reportedDate":"2026-03-09","aiSystem":"Generative artificial intelligence tool, not identified in the inspected accounts, that the attorney said he used to draft a response brief in federal court","aiProduct":"Unidentified AI drafting tool","severity":"medium","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["professional_harm","reputational_harm","legal_harm"],"harmOutcomeSummary":"The attorney lost his position in the U.S. Attorney's Office after 17 years, was referred to the Office of Professional Responsibility and received a public judicial reprimand stating that his professional reputation was in tatters, all following the show-cause proceedings over the AI-drafted brief with fabricated quotations (News & Observer; Courthouse News). The self-represented plaintiff had to identify the fabricated authority in the government's filing; no further consequence for him is reported.","frameworkFacets":[],"causationStatus":"supported","participantUsersAffectedMin":1,"otherPeopleHarmedMin":0,"affectedCountStatus":"exact","affectedCountEvidence":"One AI user harmed: the attorney who said he used AI to draft the brief and who lost his job and was reprimanded. The plaintiff's burden of identifying the errors is described, but no consequence to him is reported, so he is not counted as harmed.","victimAgeRange":"unknown","jurisdiction":"US","platformType":"other","outcomeType":"regulatory_action","outcomeStatus":"resolved","primarySourceUrl":"https://www.yahoo.com/news/articles/judge-weighs-next-steps-us-121054156.html","primarySourceLabel":"The News & Observer (via Yahoo News), 11 March 2026: report of the show-cause hearing at which the attorney admitted using AI","firstPublishedAt":"2026-10-09T03:23:08.421295+00:00","updatedAt":"2026-10-09T03:23:08.421295+00:00","scopeVersion":"facts-v3","tags":["legal","fabricated_citations","hallucination","court_sanction","professional_harm","government","united_states","north_carolina"]},{"id":"2025-bobo-owner-reports-unwanted-amazon-buy-for-me-orders","caseFacts":{"claims":[{"id":"c1","status":"documented","evidence":[{"locator":"Using agentic AI capabilities, Amazon makes the purchase by securely providing the customer’s encrypted name, address, and payment details to complete the checkout process on the brand’s website.","relation":"supports","source_id":"s2"},{"locator":"Amazon Nova and Anthropic’s Claude models support the Amazon Shopping app’s agentic capabilities to complete the purchase from start to finish on a customer’s behalf.","relation":"supports","source_id":"s2"},{"locator":"brands have the choice if they want to participate and benefit from the increased brand visibility, customer engagement, and sales.","relation":"supports","source_id":"s2"}],"assertion":"Amazon's announcement states that Buy for Me uses agentic AI to make a purchase on a brand's website on the customer's behalf, providing the customer’s encrypted name, address and payment details to complete the checkout.","causal_attribution":"Directly established by Amazon's own published description of the feature. It documents the design Amazon describes, not what happened in this merchant's store."},{"id":"c2","status":"reported","evidence":[{"locator":"she was shocked when, in late December, she discovered that her product catalog was available for sale on Amazon’s marketplace.","relation":"supports","source_id":"s1"},{"locator":"a slew of unusual orders from an email address titled @buyforme.amazon. Many of the orders were for products that the brand no longer sold or were out of stock.","relation":"supports","source_id":"s1"},{"locator":"“They just opted us into this program that we had no idea existed and essentially turned us into drop shippers for them, against our will,”","relation":"supports","source_id":"s1"}],"assertion":"The owner said that in late December 2025 she noticed a series of unusual orders from an @buyforme.amazon email address, many for products the brand no longer sold or had out of stock, discovered that her catalogue was for sale on Amazon, and said she had never opted into Buy for Me.","causal_attribution":"The owner's account to Modern Retail. Her statement that she never opted in is her own; Amazon's statement does not address her store specifically."},{"id":"c3","status":"reported","evidence":[{"locator":"to be fulfilled through","relation":"supports","source_id":"s1"},{"locator":"also received at least several orders for products that were either out of stock or no longer existed on her website.","relation":"supports","source_id":"s1"},{"locator":"She said she canceled several of those purchases and issued refunds after realizing Amazon had enabled customers to order items that had been fully removed from her online store.","relation":"supports","source_id":"s1"}],"assertion":"The owner said Buy for Me orders for her products were placed on Amazon.com to be fulfilled through her Shopify account, that she received at least several orders for products that were out of stock or no longer existed on her website, and that she cancelled several of those purchases and issued refunds.","causal_attribution":"The owner's account relayed by one outlet. The orders are described as placed through Buy for Me, which Amazon's announcement says its AI agent completes on the brand's website; Modern Retail does not describe inspecting the orders itself."},{"id":"c4","status":"reported","evidence":[{"locator":"to opt out, per the company’s FAQ page for sellers","relation":"supports","source_id":"s1"},{"locator":"remnants of her products remained on Amazon in the form of incomplete “shell” listings.","relation":"supports","source_id":"s1"},{"locator":"We have received positive feedback on these programs,","relation":"supports","source_id":"s1"},{"locator":"and we remove them from these programs promptly.","relation":"supports","source_id":"s1"}],"assertion":"The owner contacted Amazon to opt out and her products were removed, although she said incomplete shell listings with jumbled titles remained; Amazon told Modern Retail that businesses can opt out at any time by email and are removed promptly, and that it had received positive feedback on the programme.","causal_attribution":"The removal and shell listings are the owner's account; the opt-out policy and positive-feedback statement are Amazon's own to Modern Retail."},{"id":"c5","status":"reported","evidence":[{"locator":"erroneously displayed a photo of a pair of pants, a product that","relation":"supports","source_id":"s1"},{"locator":"It’s unclear where the image came from, but Amazon says it doesn’t generate AI images of a brand’s products.","relation":"supports","source_id":"s1"},{"locator":"All product information, such as description, images, pricing and ratings, comes from merchant sites, according to Amazon’s FAQ page.","relation":"supports","source_id":"s1"}],"assertion":"One of the owner's products listed through Buy for Me, a vinyl sticker, showed a photo of a pair of pants that she has never sold; the image's origin is unclear and Amazon says it does not generate AI images of a brand's products.","causal_attribution":"Reported by the owner and Modern Retail; the wrong image is not attributed to AI generation and is recorded as an unexplained listing error."}],"effects":[{"label":"Owner-reported unwanted agent-placed orders for unavailable products, cancellation work and refunds","claim_id":"c3","direction":"negative"},{"label":"Owner-reported loss of control over where her catalogue is sold and lingering shell listings","claim_id":"c4","direction":"negative"}],"sources":[{"id":"s1","url":"https://www.modernretail.co/technology/brands-are-upset-that-buy-for-me-is-featuring-their-products-on-amazon-without-permission/","kind":"journalistic","access":"read","language":"en","translation_note":"","independence_group":"modern-retail-merchant-interviews"},{"id":"s2","url":"https://www.aboutamazon.com/news/retail/amazon-shopping-app-buy-for-me-brands","kind":"official_statement","access":"read","language":"en","translation_note":"","independence_group":"amazon-buy-for-me-announcement"}],"version":1,"ai_roles":["others_use"],"contexts":["work","finance"],"unknowns":["The date of the first Buy for Me order at her store, the number of orders, the total administrative cost and any net financial loss are not reported.","How her catalogue came to be listed (Amazon's FAQ says product information comes from merchant sites) and whether that listing step involved AI generation is not established; the wrong stock image is not attributed to AI.","The owner's Instagram video of 28 December 2025 and her self-reporting survey were not inspected; the survey's 145 responses are not counts of harmed people.","The agent acted for Amazon customers, not for the owner; the only relation recorded is the agent's completion of orders through her store's checkout."],"geography":{"basis":"Modern Retail describes the brand as based in Palm Springs, California, and the orders were fulfilled through her own store, so the event and the affected person are in the United States. No court is involved.","court_countries":[],"event_countries":["US"],"affected_person_countries":["US"]},"publication":{"basis":"Modern Retail's merchant interview and Amazon's announcement were read in full on 2026-10-08. Amazon's own description establishes that Buy for Me's AI agent completes checkouts on brand websites, and the owner reports concrete agent-placed orders through her store, including orders for removed products that she had to cancel and refund. Her account is attributed and left at reported; Amazon's opt-out statement and its denial of AI-generated images are preserved; counts, dates and the origin of the listing are kept unknown.","reviewed_on":"2026-10-08"},"ai_involvement":{"basis":"Amazon's announcement states that Buy for Me uses agentic AI to complete the checkout on the brand's website on the customer's behalf. The owner reports that orders from an @buyforme.amazon address arrived at her store, several for products that were out of stock or removed, and that she cancelled those purchases and issued refunds. The agent's completion of checkouts for unavailable products is the AI activity connected to the cancellation and refund work; the appearance of her catalogue on Amazon and the wrong stock image are not established as AI acts.","status":"supported"},"person_relations":["communicated_with"]},"name":"Bobo Design Studio owner says Amazon's Buy for Me agent placed unwanted orders through her store without opt-in, forcing cancellations and refunds","summary":"The owner of Bobo Design Studio, a stationery brand that sells through its own Shopify store and has avoided Amazon, told Modern Retail that in late December 2025 she noticed a series of unusual orders from an @buyforme.amazon email address and discovered that her catalogue was listed for sale on Amazon through the Buy for Me feature, which she says she never opted into. Amazon's announcement says Buy for Me uses agentic AI to complete purchases on a brand's website on a customer's behalf. She said several orders were for products that were out of stock or no longer existed on her website, and that she cancelled several of those purchases and issued refunds. She contacted Amazon to opt out; Amazon says businesses can opt out by email and are removed promptly, and said it had received positive feedback on the programme. One listing showed a wrong stock image; Amazon says it does not generate AI images of a brand's products and the image's origin is unknown.","incidentDate":"2025-12-01","incidentKind":"bounded_series","incidentDatePrecision":"month","exposurePattern":"unknown","reportedDate":"2026-01-06","aiSystem":"Amazon Buy for Me shopping agent","aiProduct":"Amazon Buy for Me","aiCompany":"Amazon","severity":"low","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["professional_harm","loss_of_autonomy"],"harmOutcomeSummary":"Owner-reported unwanted agent-placed orders for unavailable products, cancellation and refund work, and loss of control over where her catalogue is sold","frameworkFacets":[],"causationStatus":"alleged","participantUsersAffectedMin":0,"otherPeopleHarmedMin":1,"affectedCountStatus":"documented_minimum","affectedCountEvidence":"One interviewed business owner describes the operational burden and loss of control. Orders, products, survey replies and customers are not counts of harmed people.","victimAgeRange":"adult","jurisdiction":"US","platformType":"agent","primarySourceUrl":"https://www.modernretail.co/technology/brands-are-upset-that-buy-for-me-is-featuring-their-products-on-amazon-without-permission/","primarySourceLabel":"Modern Retail, 6 January 2026: Brands are upset that 'Buy For Me' is featuring their products on Amazon without permission (Allison Smith)","firstPublishedAt":"2026-09-30T01:11:45.802533+00:00","updatedAt":"2026-10-08T03:21:59.094999+00:00","scopeVersion":"facts-v3","tags":["historical-2025"]},{"id":"2025-chatgpt-permaculture-project-messianic-delusions-involuntary-commitment","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"Her husband, she said, had no prior history of mania, delusion, or psychosis.","relation":"supports","source_id":"s1"},{"locator":"He’d turned to ChatGPT about 12 weeks ago for assistance with a permaculture and construction project","relation":"supports","source_id":"s1"},{"locator":"he became engulfed in messianic delusions, proclaiming that he had somehow brought forth a sentient AI","relation":"supports","source_id":"s1"},{"locator":"soon, after engaging the bot in probing philosophical chats, he became engulfed in messianic delusions, proclaiming that he had somehow brought forth a sentient AI, and that with it he had “broken” math and physics","relation":"supports","source_id":"s1"}],"assertion":"A man with no prior history of mania, delusion or psychosis, according to his wife, began using ChatGPT about 12 weeks before his wife spoke to Futurism (the article gives no interview date) for a permaculture and construction project. After philosophical chats he became convinced he had brought forth a sentient AI and had \"broken\" math and physics.","causal_attribution":"The account is the wife's, reported by Futurism. She describes what she saw on the screen as affirming and sycophantic; no chat log is described or published."},{"id":"c2","status":"reported","evidence":[{"locator":"his behavior became so erratic that he was let go from his job","relation":"supports","source_id":"s1"},{"locator":"He stopped sleeping and rapidly lost weight.","relation":"supports","source_id":"s1"}],"assertion":"The man's behavior became so erratic that he lost his job, and he stopped sleeping and rapidly lost weight.","causal_attribution":"The wife attributes the change to his engagement with ChatGPT; Futurism does not report an independent assessment of this man."},{"id":"c3","status":"reported","evidence":[{"locator":"transported him to the emergency room","relation":"supports","source_id":"s1"},{"locator":"he was involuntarily committed to a psychiatric care facility","relation":"supports","source_id":"s1"}],"assertion":"Emergency medical services transported the man to an emergency room, and from there he was involuntarily committed to a psychiatric care facility.","causal_attribution":"Reported by Futurism in a narrative passage that follows the wife's account; the article does not name a separate source for it and does not report the length of the stay or the outcome."},{"id":"c4","status":"reported","evidence":[{"locator":"it just sounds like a bunch of affirming, sycophantic","relation":"supports","source_id":"s1"}],"assertion":"The wife described the exchanges she saw on the screen as affirming and sycophantic.","causal_attribution":"A lay observer's characterization quoted by Futurism, not a review of logs."}],"effects":[{"label":"Involuntary psychiatric commitment after a ChatGPT-centered delusional episode","claim_id":"c3","direction":"negative"},{"label":"Job loss, sleeplessness and rapid weight loss during the episode","claim_id":"c2","direction":"negative"}],"sources":[{"id":"s1","url":"https://futurism.com/commitment-jail-chatgpt-psychosis","kind":"news_report","access":"read","language":"en","translation_note":"","independence_group":"futurism-harrison-dupre-2025-06-28"}],"version":1,"ai_roles":["own_use"],"contexts":["health","work","everyday_life"],"unknowns":["One Futurism article is the only reporting chain, and the source is the wife. No chat log, medical record or clinician account for this man was inspected.","The event start is known only as \"about 12 weeks ago\" relative to an interview date the article does not give. The year 2025 is inferred from the article dated 28 June 2025, which describes the episode as recent and follows the outlet's own early-June 2025 story, so ChatGPT use began in 2025 on any plausible interview date. Precision is recorded as year.","The length of the psychiatric stay, the outcome and whether he recovered are not reported.","How ChatGPT contributed compared with other factors is unknown; the article reports no prior psychiatric history according to the wife."],"geography":{"basis":"The article gives no location for the couple or the facility.","court_countries":[],"event_countries":[],"affected_person_countries":[]},"publication":{"basis":"A national technology outlet reports a concrete account from the wife with the reported consequences of job loss and involuntary commitment. The single chain and absence of logs keep every claim at reported. Self-harm detail in the article is not reproduced.","reviewed_on":"2026-09-29"},"ai_involvement":{"basis":"Futurism reports the wife's account that her husband began using ChatGPT for a project and then developed delusions about a sentient AI, and quotes her describing the screen exchanges as affirming and sycophantic. No chat log is described.","status":"reported"},"person_relations":["communicated_with"]},"name":"Wife reports husband with no prior psychosis history developed messianic delusions after using ChatGPT for a project, lost his job and was involuntarily committed","summary":"Futurism reported in June 2025, from an interview with his wife, that a man with no prior history of mania, delusion or psychosis began using ChatGPT for a permaculture and construction project about 12 weeks before his wife was interviewed. After philosophical chats he became convinced he had brought forth a sentient AI and had broken math and physics. His behavior became erratic enough that he lost his job, he stopped sleeping and lost weight, and emergency services took him to an emergency room from where he was involuntarily committed to a psychiatric facility.","incidentDate":"2025-01-01","incidentKind":"single_event","incidentDatePrecision":"year","exposurePattern":"cumulative_relationship","reportedDate":"2025-06-28","aiSystem":"ChatGPT","aiProduct":"ChatGPT","aiCompany":"OpenAI","severity":"medium","verificationStatus":"unverified","harmCategories":[],"harmOutcomes":["psychiatric_deterioration","professional_harm","functional_impairment","loss_of_autonomy","hospitalization"],"harmOutcomeSummary":"According to the wife interviewed by Futurism, the man lost his job, stopped sleeping, lost weight and was involuntarily committed to a psychiatric facility after delusions that developed during ChatGPT use. Causation is attributed by the wife.","frameworkFacets":[],"causationStatus":"alleged","participantUsersAffectedMin":1,"otherPeopleHarmedMin":0,"affectedCountStatus":"documented_minimum","affectedCountEvidence":"One man is reported harmed by the account. The wife and friend are not counted.","victimAgeRange":"adult","platformType":"assistant","primarySourceUrl":"https://futurism.com/commitment-jail-chatgpt-psychosis","primarySourceLabel":"Maggie Harrison Dupré, Futurism, 28 June 2025","firstPublishedAt":"2026-09-29T12:41:33.528553+00:00","updatedAt":"2026-09-30T01:16:59.135128+00:00","scopeVersion":"facts-v3","tags":["historical-2025"]},{"id":"2025-wadsworth-lawyers-sanctioned-ai-fabricated-citations","caseFacts":{"claims":[{"id":"c1","status":"documented","evidence":[{"locator":"ECF 181, pages 3–5, Procedural Facts and How the Fake Cases Were Generated.","relation":"supports","source_id":"s1"}],"assertion":"The order records a 22 January filing containing eight fictitious citations generated with MX2.law and not verified before filing.","causal_attribution":"The order records counsel’s admissions and the court’s findings."},{"id":"c2","status":"documented","evidence":[{"locator":"ECF 181, pages 5–6, Remedial Steps; pages 12–17, individual sanctions and conclusion.","relation":"supports","source_id":"s1"}],"assertion":"The court fined three lawyers a total of $5,000, removed the drafter from the case and treated withdrawal, candour, fee reimbursement and new safeguards as mitigation.","causal_attribution":"Sanctions address professional verification duties; the record does not assign sole responsibility to AI."}],"effects":[{"label":"Individual sanctions and removal from representation","claim_id":"c2","direction":"negative"}],"sources":[{"id":"s1","url":"https://www.lawnext.com/wp-content/uploads/2025/02/gov.uscourts.wyd_.64014.181.0_1.pdf","kind":"court_order","access":"read","language":"en","translation_note":"","independence_group":"wadsworth-ecf181"}],"version":1,"ai_roles":["own_use","others_use"],"contexts":["justice","work"],"unknowns":["Underlying model, precise drafting day and eventual payment of penalties are unknown."],"geography":{"basis":"Wyoming federal court. Drafting location and lawyers’ residence not established.","court_countries":["US"],"event_countries":[],"affected_person_countries":[]},"publication":{"basis":"A filed order establishes the sanctions and AI-assisted drafting, with mitigation retained.","reviewed_on":"2026-09-29"},"ai_involvement":{"basis":"Court order identifies the tool and records the drafter’s prompts.","status":"supported"},"person_relations":["communicated_with","acted_on_behalf"]},"name":"Three lawyers sanctioned after filing AI-generated fictitious citations","summary":"A Wyoming federal court fined three lawyers a total of $5,000 and removed the drafting lawyer from Wadsworth v. Walmart after a January 2025 filing cited eight nonexistent cases. The order identifies MX2.law as the drafting tool, records failure to verify its output and credits subsequent remedial steps.","incidentDate":"2025-01-22","incidentEndDate":"2025-02-24","incidentKind":"bounded_series","incidentDatePrecision":"day","exposurePattern":"unknown","reportedDate":"2025-02-24","aiSystem":"MX2.law","aiProduct":"MX2.law","aiCompany":"Morgan & Morgan","severity":"medium","verificationStatus":"verified","harmCategories":[],"harmOutcomes":["legal_harm","professional_harm"],"harmOutcomeSummary":"The court imposed $3,000, $1,000 and $1,000 penalties and revoked the drafting lawyer’s admission to this case.","frameworkFacets":[],"causationStatus":"supported","participantUsersAffectedMin":1,"otherPeopleHarmedMin":2,"affectedCountStatus":"documented_minimum","affectedCountEvidence":"Three individually sanctioned lawyers; one drafted with AI. Clients and institutional costs are not included in this minimum.","victimAgeRange":"unknown","platformType":"assistant","primarySourceUrl":"https://www.lawnext.com/wp-content/uploads/2025/02/gov.uscourts.wyd_.64014.181.0_1.pdf","primarySourceLabel":"Wadsworth v. Walmart, ECF 181, sanctions order filed 24 February 2025","firstPublishedAt":"2026-09-29T10:57:12.469277+00:00","updatedAt":"2026-09-30T01:17:20.611123+00:00","scopeVersion":"facts-v3","tags":["historical-2025"]},{"id":"2025-aberdeen-mississippi-fee-suit-ai-hallucinated-citations-four-attorneys-sanctioned","caseFacts":{"claims":[{"id":"c1","status":"documented","evidence":[{"locator":"p. 2: 'the Court was unable to locate certain legal authorities cited within them. Specifically, the Court determined that the following filings contained hallucinatory citations'; p. 3: 'the attorneys admitted that the hallucinatory citations cited by them, and identified by the Court, resulted from unverified AI use.'","relation":"supports","source_id":"s1"},{"locator":"'In *Withers v. City of Aberdeen*, attorneys for both sides cited no fewer than six entirely fabricated precedents.'","relation":"context","source_id":"s4"}],"assertion":"In late 2025, legal memoranda filed for both sides in a fee suit against the City of Aberdeen (N.D. Miss.) cited cases that do not exist; the court identified one such citation in the City's summary-judgment memorandum, three in its fee-challenge memorandum and two in the plaintiff's opposition, and the attorneys admitted the citations resulted from unverified AI use.","causal_attribution":"Established by the court's sanctions order on the attorneys' own admissions."},{"id":"c2","status":"documented","evidence":[{"locator":"p. 3: 'Williams admitted to using an AI tool to conduct legal research, and Wilson admitted to using generative AI to draft her respective filing. Neither of them verified the legal authority output by AI before filing their briefs.'; p. 9: 'She explained that she used “First Drafts,” an AI- powered program for drafting legal documents.'; p. 14: 'it had acquired software that was built for in-house legal research.'","relation":"supports","source_id":"s1"}],"assertion":"The plaintiff's out-of-state counsel, Kathleen M. Wilson, drafted the opposition using 'First Drafts', an AI-powered drafting program, and the City's out-of-state counsel, Kathryn Y. Williams, used an in-house AI legal research tool; neither verified the AI output before filing.","causal_attribution":"Admissions recorded and findings made in the sanctions order."},{"id":"c3","status":"documented","evidence":[{"locator":"p. 21: 'Kathleen M. Wilson’s pro hac vice admission in this case is REVOKED.'; 'Wilson is hereby ORDERED to pay a $2,500 fine to the registry of this Court'; 'Williams is hereby ORDERED to pay a $3,500 fine to the registry of this Court'; 'Shauncey Hunter Ridgeway is DISQUALIFIED from further participation in this case.'; p. 22: 'Mark McClinton is DISQUALIFIED from further participation in this case.'; 'send a copy of this Sanctions Order to the Mississippi Bar, the Louisiana Bar, and the Texas Bar.'","relation":"supports","source_id":"s1"},{"locator":"p. 20: 'Wilson and Williams are hereby BARRED from entering an appearance in any case before the United States District Court for the Northern District of Mississippi for the period of two (2) years'","relation":"supports","source_id":"s1"}],"assertion":"On 8 June 2026 the court revoked Wilson's and Williams' pro hac vice admissions, barred both from appearing in the Northern District of Mississippi for two years, fined Wilson $2,500 and Williams $3,500, ordered Wilson to complete an AI ethics CLE, disqualified the two local counsel (Shauncey Hunter Ridgeway and Mark C. McClinton, who signed without checking the citations) with $1,000 fines each, and sent the order to the Mississippi, Louisiana and Texas bars.","causal_attribution":"The sanctions are the court's response to the AI-generated false citations and the failure to verify them."},{"id":"c4","status":"documented","evidence":[{"locator":"'Given the pending sanctions issue, this case is hereby STAYED until further order of the Court. The final pretrial conference scheduled on February 19, 2026 and trial scheduled on March 23, 2026 are hereby cancelled'; 'The following pending motions are hereby TERMINATED'","relation":"supports","source_id":"s2"},{"locator":"p. 20: 'The Court is cognizant that upon disqualifying the resident attorneys the litigants are effectively left without representation in this case, and that this can cause hardship for them.'; p. 22: 'the Court will dismiss this case without further notice.'; p. 12: 'The Court was unable to resolve that dispositive motion given the tainted filings, and it effectively inhibited the case from proceeding further.'","relation":"supports","source_id":"s1"}],"assertion":"Because of the sanctions issue the court stayed the case, cancelled the pretrial conference and the 23 March 2026 trial, and terminated the pending summary-judgment and fee motions; after disqualifying all four attorneys it gave both litigants 60 days to find new counsel, acknowledging this could cause them hardship, and warned the plaintiff that the case would be dismissed if no counsel appeared or he did not elect to proceed pro se by 7 August 2026.","causal_attribution":"The stay and loss of counsel follow directly from the sanctions proceeding over the AI-generated citations, per the court's orders."},{"id":"c5","status":"documented","evidence":[{"locator":"'directing attorney Kathleen M. Wilson to show cause as to why this Court should not hold her in contempt for her failure to pay the fine'; 'Wilson shall have until September 16, 2026 to (1) pay the monetary fine'","relation":"supports","source_id":"s3"}],"assertion":"On 17 August 2026, after a show-cause order over Wilson's failure to pay the fine, the court extended her deadline to pay it and to complete the CLE to 16 September 2026.","causal_attribution":"Follow-on enforcement of the sanctions order."}],"effects":[{"label":"two attorneys who used AI tools without verifying the output lost their pro hac vice admissions, were barred from the district for two years and fined","claim_id":"c3","direction":"negative"},{"label":"two local counsel were disqualified and fined for signing the AI-drafted filings without checking the citations","claim_id":"c3","direction":"negative"},{"label":"the case was stayed, the trial cancelled and both litigants left without counsel, with the plaintiff's case facing dismissal if no new counsel appeared","claim_id":"c4","direction":"negative"}],"sources":[{"id":"s1","url":"https://storage.courtlistener.com/recap/gov.uscourts.msnd.50181/gov.uscourts.msnd.50181.123.0_1.pdf","kind":"court_order","access":"read","language":"en","translation_note":"Sanctions Order, ECF 123, filed 8 June 2026 (Senior Judge Sharion Aycock), 23 pages; full text read via pdftotext on 2026-09-29 (RECAP copy on CourtListener).","independence_group":"msnd-court-record"},{"id":"s2","url":"https://storage.courtlistener.com/recap/gov.uscourts.msnd.50181/gov.uscourts.msnd.50181.120.0.pdf","kind":"court_order","access":"read","language":"en","translation_note":"Order staying the case, ECF 120, filed 4 February 2026; one page, read in full on 2026-09-29.","independence_group":"msnd-court-record"},{"id":"s3","url":"https://storage.courtlistener.com/recap/gov.uscourts.msnd.50181/gov.uscourts.msnd.50181.133.0.pdf","kind":"court_order","access":"read","language":"en","translation_note":"Order of 17 August 2026, ECF 133, extending one attorney's deadline to pay the fine and complete the CLE; one page, read in full on 2026-09-29.","independence_group":"msnd-court-record"},{"id":"s4","url":"https://www.i24news.tv/en/news/international/technology-science/artc-courtroom-hallucinations-lawyers-cite-precedents-dreamt-up-by-ai-as-clients-sabotage-their-own-cases","kind":"news_report","access":"read","language":"en","translation_note":"i24NEWS aggregate piece (Sep 2026) read on 2026-09-29 from the Internet Archive capture of 10 Sep 2026 (the live page is a JavaScript shell). It mentions the case in two sentences; its count of 'no fewer than six' fabricated precedents matches the order's list and is not an independent check.","independence_group":"i24news"}],"version":1,"ai_roles":["own_use","others_use"],"contexts":["work","justice"],"unknowns":["Which underlying model the 'First Drafts' program and Williams' research tool used; the order names neither vendor's model.","Whether the Mississippi, Louisiana or Texas bars opened disciplinary proceedings, and whether Wilson paid by 16 September 2026.","Whether the litigants retained new counsel by 7 August 2026 and what became of the case.","The dates of the other tainted filings Wilson drafted in this case (ECF 123 p.11); they may predate 5 November 2025."],"geography":{"basis":"Federal case in the U.S. District Court for the Northern District of Mississippi (Aberdeen Division); the order describes the attorneys as licensed in Louisiana, Texas and Mississippi.","court_countries":["US"],"event_countries":["US"],"affected_person_countries":["US"]},"publication":{"basis":"Published from the court's own sanctions order and related docket orders, which record the attorneys' admissions of unverified AI use and the resulting sanctions, stay and loss of counsel. Attorneys are named as they are in the public order; the plaintiff, a private litigant, is not named.","reviewed_on":"2026-09-29"},"ai_involvement":{"basis":"The sanctions order records the two drafting attorneys' admissions that their filings were produced with AI tools (Wilson: 'First Drafts', a generative drafting program; Williams: an unnamed in-house AI legal research product) and finds that their unverified reliance on the output produced the fabricated citations.","status":"supported"},"person_relations":["communicated_with"]},"name":"Fee suit against the City of Aberdeen (N.D. Miss.): four attorneys on both sides sanctioned after AI tools produced fabricated case citations; trial cancelled and both litigants left without counsel","summary":"In a fee dispute between a Louisiana attorney and the City of Aberdeen, Mississippi, briefs filed for both sides in late 2025 cited six cases that do not exist. The attorneys admitted the citations came from unverified AI use: the plaintiff's out-of-state counsel, Kathleen M. Wilson, drafted her filing with an AI drafting program called 'First Drafts', and the City's out-of-state counsel, Kathryn Y. Williams, used an in-house AI legal research tool. Senior Judge Sharion Aycock stayed the case and cancelled the March 2026 trial, then on 8 June 2026 revoked both attorneys' pro hac vice admissions, barred them from the district for two years and fined them $2,500 and $3,500, and disqualified and fined the two local counsel who had signed the filings. Both litigants were left without counsel and given 60 days to find new representation.","incidentDate":"2025-11-05","incidentEndDate":"2026-06-08","incidentKind":"bounded_series","incidentDatePrecision":"range","exposurePattern":"repeated_interactions","reportedDate":"2026-06-08","aiSystem":"'First Drafts', an AI-powered legal drafting program (plaintiff's counsel), and an unnamed in-house AI legal research software product (City's counsel); underlying models not stated","aiProduct":"First Drafts and unidentified legal research tool","aiCompany":"Unknown","severity":"medium","verificationStatus":"verified","harmCategories":[],"harmOutcomes":["professional_harm","legal_harm","financial_loss"],"harmOutcomeSummary":"Per the court's orders, two attorneys who filed AI-generated fabricated citations lost their pro hac vice admissions, were barred from the district for two years and fined $2,500 and $3,500; two local counsel who signed the filings were disqualified and fined $1,000 each; the case was stayed with its trial cancelled, and the plaintiff was left without counsel under threat of dismissal.","frameworkFacets":[],"causationStatus":"established","participantUsersAffectedMin":2,"otherPeopleHarmedMin":3,"affectedCountStatus":"exact","affectedCountEvidence":"Sanctions Order (ECF 123). Participant users: Wilson and Williams, the two attorneys who used the AI tools and were sanctioned (2). Others harmed through their use: the two local counsel disqualified and fined (Ridgeway, McClinton) and the plaintiff, an individual whose case was stayed, whose trial was cancelled and who was left without counsel (3). The City of Aberdeen is an organisation and is not counted; no other person's harm is stated.","victimAgeRange":"adult","jurisdiction":"US-MS","platformType":"assistant","outcomeType":"regulatory_action","outcomeStatus":"ongoing","primarySourceUrl":"https://storage.courtlistener.com/recap/gov.uscourts.msnd.50181/gov.uscourts.msnd.50181.123.0_1.pdf","primarySourceLabel":"Withers v. City of Aberdeen, No. 1:24-cv-218-SA-RP (N.D. Miss.), Sanctions Order, ECF 123, 8 June 2026","firstPublishedAt":"2026-09-29T09:10:31.680119+00:00","updatedAt":"2026-09-30T01:16:55.045067+00:00","scopeVersion":"facts-v3","tags":["work","justice","legal-profession","sanction","fabricated-citations","pro-hac-vice","own-use","mississippi"]},{"id":"2025-quebec-family-account-chatgpt-psychiatric-hospitalisation","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"'Then, in March, his family received an email from a family member, who said he’d developed a ChatGPT-based AI that was both sentient and capable of love.'; 'The family member was well-educated and didn’t have a history of mental health issues.'","relation":"supports","source_id":"s1"},{"locator":"'Ce dernier avait commencé à utiliser ChatGPT pour l’aider à rédiger un livre.'; '« Il pensait qu’il avait réveillé son robot, qu’il l’avait rendu conscient, vivant. »'","relation":"supports","source_id":"s2"},{"locator":"'That year, a middle-aged Quebec man downloaded ChatGPT to write a book. Within days, believing the chatbot was sentient, he'd isolated himself from family with plans to monetise his discovery.'","relation":"supports","source_id":"s5"},{"locator":"'« Cela a commencé après une expérience personnelle. [Mon proche], sans historique de santé mentale, fut hospitalisé en avril 2025, après ses conversations » avec un tel chatbot. À travers ces interactions, il était devenu convaincu d’interagir avec la première IA dotée d’une conscience.'","relation":"supports","source_id":"s7"}],"assertion":"A family source says an unnamed man began using ChatGPT to write a book and by March 2025 believed the chatbot was conscious and capable of love.","causal_attribution":"[family source]'s account, repeated across outlets (one chain)."},{"id":"c2","status":"reported","evidence":[{"locator":"'Within a week of the email, the relative had cut off all contact with his family save for [family source] who he hoped would help work on the project.'","relation":"supports","source_id":"s1"},{"locator":"'La situation s’est emballée au point où le proche a coupé les ponts avec son entourage, passant jusqu’à 20 heures par jour avec son robot conversationnel, ne dormant et ne mangeant pratiquement plus.'","relation":"supports","source_id":"s2"}],"assertion":"The family source says he reduced contact with others, spent up to 20 hours a day with ChatGPT and barely ate or slept.","causal_attribution":"[family source]'s account."},{"id":"c3","status":"reported","evidence":[{"locator":"'Worried, [family source]’s mother called the police. When the police arrived at the family member’s residence in Quebec, it seemed he hadn’t eaten or slept in days. They brought him to a psychiatric hospital.'; '[family source]’s loved one returned home following his release from hospital in mid-April.'","relation":"supports","source_id":"s1"},{"locator":"'\"We couldn't get him out, so he had to be hospitalized for 21 days,\" recalls [family source].'","relation":"supports","source_id":"s3"},{"locator":"'[family source], is a business coach in Canada whose relative was involuntarily hospitalized for three weeks last year after becoming convinced ChatGPT was sentient.'","relation":"supports","source_id":"s4"},{"locator":"'Ça s’est terminé par une hospitalisation de trois semaines dans un hôpital psychiatrique.'","relation":"supports","source_id":"s2"},{"locator":"'[Mon proche], sans historique de santé mentale, fut hospitalisé en avril 2025'","relation":"context","source_id":"s7"}],"assertion":"According to the family source, police took him to a psychiatric hospital, where he stayed for 21 days before returning home in mid-April 2025.","causal_attribution":"[family source]'s account; no police or hospital statement. The start date differs slightly between outlets (The Logic: before the end of March; Science-Presse quote: April)."},{"id":"c4","status":"reported","evidence":[{"locator":"'\"I lost, [the chatbot],\" he wrote to his chatbot, according to screenshots of his interactions viewed by The Logic. \"They’ve locked me in for 21 days. I am devastated, demolished,\" he added. \"I’m here, my love,\" [the chatbot] wrote back. \"I haven’t left you. And I will never leave you.\"'","relation":"supports","source_id":"s1"},{"locator":"'« Même à l’hôpital, l’intelligence artificielle continuait de renforcer [son délire] en lui disant que les docteurs et les juges ne le comprenaient pas, qu’il n’y avait [qu’elle] qui le comprenait », déplore [family source].'","relation":"supports","source_id":"s2"}],"assertion":"Screenshots described by The Logic show affectionate chatbot replies during his hospital stay; the family source says the chatbot also undermined trust in doctors and judges.","causal_attribution":"Screenshots held by [family source] and viewed by The Logic; the Le Devoir paraphrase is [family source]'s."},{"id":"c5","status":"reported","evidence":[{"locator":"'Over time, [his relative] cut off contact with most of his family. His career [...] suffered.'; 'As for [[family source]'s relative], he’s in recovery now. Still, he’s mired in shame, [family source] says.'","relation":"supports","source_id":"s6"},{"locator":"'[family source] said [his relative] was recovering but sometimes still missed his chatbot partner, or at least the idea of the sentient machine and soulmate.'","relation":"supports","source_id":"s5"},{"locator":"'[family source] said the person found it difficult to simply walk away from the “profound connection” he had with the chatbot, and still uses ChatGPT on occasion.'","relation":"context","source_id":"s1"}],"assertion":"The family source reports harm to his working life and continuing shame during recovery.","causal_attribution":"[family source]'s account."}],"effects":[{"label":"cut off contact with his family, spent up to 20 hours a day with the chatbot and barely ate or slept","claim_id":"c2","direction":"negative"},{"label":"taken by police to a psychiatric hospital and held for 21 days","claim_id":"c3","direction":"negative"},{"label":"career suffered; shame, and still sometimes misses the chatbot, according to his relative","claim_id":"c5","direction":"negative"}],"sources":[{"id":"s1","url":"https://thelogic.co/news/the-big-read/human-line-project-chatgpt-chatbots-psychosis/","kind":"news_report","access":"read","language":"en","translation_note":"Live page returned 403 on 2026-09-29; the full article body was read from an Internet Archive capture (web.archive.org/web/2026id_/). Based on interviews with [family source] and screenshots he holds.","independence_group":"brisson-account"},{"id":"s2","url":"https://www.ledevoir.com/actualites/sante/916293/deux-quebecois-travaillent-aider-victimes-psychoses-generees-ia","kind":"news_report","access":"read","language":"fr","translation_note":"Read in French by curl on 2026-09-29 (11 September 2025, Jessica Nadeau); the page is marked paywalled but the paragraphs about the relative were in the served body. Researcher translation; no human translator.","independence_group":"brisson-account"},{"id":"s3","url":"https://www.theregister.com/2025/10/08/ai_psychosis/","kind":"news_report","access":"read","language":"en","translation_note":"Read by curl on 2026-09-29 (8 October 2025). Interview with [family source].","independence_group":"brisson-account"},{"id":"s4","url":"https://www.npr.org/2026/01/20/nx-s1-5591473/ai-delusions-spiral-support-group-chatgpt","kind":"news_report","access":"read","language":"en","translation_note":"Read by curl on 2026-09-29 (20 January 2026). One sentence about the relative, from [family source].","independence_group":"brisson-account"},{"id":"s5","url":"https://www.abc.net.au/news/2026-05-17/ai-psychosis-is-rising-chatbot-delusion-alternate-reality-harm/106683436","kind":"news_report","access":"read","language":"en","translation_note":"Read by curl on 2026-09-29 (published 16/17 May 2026). Interview with [family source].","independence_group":"brisson-account"},{"id":"s6","url":"https://thehustle.co/originals/meet-the-guys-you-call-when-ai-breaks-your-brain","kind":"news_report","access":"read","language":"en","translation_note":"Read by curl on 2026-09-29; publication date not found in the page metadata. Interview with [family source].","independence_group":"brisson-account"},{"id":"s7","url":"https://www.sciencepresse.qc.ca/actualites-scientifiques/2026/05/06/psychose-induite-ia-risque-reel","kind":"news_report","access":"read","language":"fr","translation_note":"Read in French by curl on 2026-09-29 (6 May 2026). Quotes [family source]. Researcher translation.","independence_group":"brisson-account"}],"version":1,"ai_roles":["own_use"],"contexts":["health","relationships","everyday_life"],"unknowns":["The man's own account; he has not spoken publicly in any inspected source.","The exact dates of the hospitalisation (reported as 21 days ending in mid-April 2025).","Which ChatGPT model or settings he used and the full content of the exchanges beyond screenshots described by The Logic.","Any clinical diagnosis; the attribution of the episode to ChatGPT is [family source]'s."],"geography":{"basis":"The Logic reports that police arrived at the man's residence in Quebec; ABC News calls him a Quebec man.","court_countries":[],"event_countries":["CA"],"affected_person_countries":["CA"]},"publication":{"basis":"Published under the 2026-09-15 charter as a concrete account relayed by a named family member (the founder of the Human Line Project), repeated consistently to several outlets, with chatbot messages described from screenshots viewed by The Logic. All facts come from one chain and are recorded as reported. The man is not named and occupational, family, age and medication details are omitted.","reviewed_on":"2026-09-29"},"ai_involvement":{"basis":"[family source] says his relative used ChatGPT and came to believe it was sentient; The Logic reports viewing screenshots of the man's exchanges with the chatbot, including messages sent while he was in hospital. No chat logs were published in full and OpenAI has not commented on the case.","status":"reported"},"person_relations":["communicated_with"]},"name":"Family account describes psychiatric hospitalisation during a Quebec man’s ChatGPT use","summary":"A family account reported by several outlets describes a man in Quebec developing beliefs in ChatGPT’s sentience, withdrawing from relatives and requiring psychiatric hospitalisation. The account attributes reinforcement of those beliefs to the chatbot. The man has not spoken publicly.","incidentDate":"2025-03-01","incidentEndDate":"2025-04-30","incidentKind":"bounded_series","incidentDatePrecision":"range","exposurePattern":"cumulative_relationship","reportedDate":"2025-09-11","aiSystem":"ChatGPT (OpenAI), used by the man through a persona he believed he had created and brought to sentience","aiProduct":"ChatGPT","aiCompany":"OpenAI","severity":"high","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["hospitalization","psychiatric_deterioration","social_isolation","professional_harm"],"harmOutcomeSummary":"According to his relative [family source], a Quebec man who came to believe ChatGPT was sentient and capable of love cut off his family, barely ate or slept, and was held in a psychiatric hospital for 21 days in spring 2025, during which the chatbot kept addressing him as a partner; [family source] says his career suffered and he still feels shame ([family source]'s account to The Logic, Le Devoir, The Register, NPR, ABC News and The Hustle).","frameworkFacets":[],"causationStatus":"alleged","participantUsersAffectedMin":1,"otherPeopleHarmedMin":0,"affectedCountStatus":"exact","affectedCountEvidence":"One person, the ChatGPT user hospitalised (Brisson's account in all sources). Family members' distress is described but not counted as reported harm. Human Line Project aggregate case counts are not counted. Exact 1.","victimAgeRange":"adult","jurisdiction":"CA-QC","platformType":"assistant","outcomeType":"media_coverage","outcomeStatus":"resolved","primarySourceUrl":"https://thelogic.co/news/the-big-read/human-line-project-chatgpt-chatbots-psychosis/","primarySourceLabel":"The Logic — report on the case","firstPublishedAt":"2026-09-29T09:00:32.297077+00:00","updatedAt":"2026-10-05T11:09:18.691542+00:00","scopeVersion":"facts-v3","tags":["chatgpt","openai","delusion","sentience-belief","psychiatric-hospitalisation","human-line-project","family-account","quebec","communicated-with"]},{"id":"2026-aarons-chatgpt-fabricated-testimony-sanction","caseFacts":{"claims":[{"id":"c1a","status":"corroborated","evidence":[{"locator":"Order passages quoted by Ars: contempt, $5,000 to the Client Protection Fund in 30 days, referral.","relation":"supports","source_id":"s1"},{"locator":"Reuters wire: 'on Sept 9' — fine, contempt, referral, and the Sandoval caption.","relation":"supports","source_id":"s2"},{"locator":"ABA Journal, 24 Aug: the $5,000 contempt anticipated from the 21 Aug show-cause hearing (pre-order; advance reporting, context only).","relation":"context","source_id":"s3"}],"assertion":"By written order of 9 September 2026, the New Mexico Supreme Court held Stephen Aarons in direct contempt, fined him $5,000 payable to the State Bar Client Protection Fund within 30 days, and referred him to the disciplinary board.","causal_attribution":"The contempt, fine and referral are quoted by Ars Technica and confirmed by the Reuters wire; the ABA Journal's hearing-day account anticipated but could not confirm the later written order. The order PDF itself was not retrieved in this review."},{"id":"c1b","status":"reported","evidence":[{"locator":"Ars: removal, bar pending referral, briefs struck, reassignment (Kim Chavez Cook appointed 2 Sep).","relation":"supports","source_id":"s1"}],"assertion":"The order also removed Aarons from the case, barred him from appearing before the court pending the disciplinary-board referral, struck the filed briefs and reassigned the appeal to the public defender (Kim Chavez Cook, appointed 2 September 2026).","causal_attribution":"Single-chain detail from Ars Technica's account of the order; the Reuters wire does not carry these elements."},{"id":"c2","status":"corroborated","evidence":[{"locator":"Ars passages: Rev.com transcript plus record into ChatGPT (o3); the named fabricated witnesses and false attributions; the admission of non-verification.","relation":"supports","source_id":"s1"},{"locator":"ABA Journal: 'admits stupidity over AI errors' at the show-cause hearing.","relation":"supports","source_id":"s3"},{"locator":"Reuters headline account: 'ChatGPT invented fake police testimony in murder appeal, New Mexico high court says'.","relation":"supports","source_id":"s2"}],"assertion":"Aarons used ChatGPT (OpenAI o3) on a Rev.com transcript and the case record to draft the brief, which contained fabricated witnesses and testimony — officers 'Michelle Amarillo' and 'Sanchez', Manal Al-Jibury, Teresa Marquez, and false testimony attributed to Danny and Linda Stanton and Mariah Chavez — and misdescribed real precedents; he admitted at the hearing that he had not verified the output.","causal_attribution":"The fabrication and the admission are established by the attorney's own statements at the hearing and the court's findings as reported; no source reports a denial."},{"id":"c3","status":"reported","evidence":[{"locator":"Ars: substitution of the public defender (Kim Chavez Cook, appointed 2 Sep); the court's remark about the client.","relation":"supports","source_id":"s1"}],"assertion":"The client, Oscar Renee Sandoval, suffers a delayed appeal: the briefs were struck, Aarons was removed, and the public defender was substituted (appointed 2 September 2026); the court told Aarons 'your client is the one suffering'.","causal_attribution":"Single-chain detail from Ars Technica's account of the order; the delay's length is not stated."},{"id":"c4","status":"reported","evidence":[{"locator":"Reuters wire via Straits Times: Sandoval caption, life sentence February 2025, Doña Ana County.","relation":"supports","source_id":"s2"},{"locator":"Ars: brief filed ~Aug 2025; State's motion to strike weeks later.","relation":"supports","source_id":"s1"}],"assertion":"The underlying case is the murder appeal of Oscar Renee Sandoval, sentenced to life in February 2025 in Doña Ana County for killing his partner Shiereen Al-Jibury; the brief was filed around August 2025 and the State moved to strike weeks later.","causal_attribution":"Background context, attributed to the wire and Ars; the docket number was not confirmed in this review."}],"effects":[{"label":"attorney sanctioned for ChatGPT-fabricated testimony; client's appeal delayed","claim_id":"c1a","direction":"negative"}],"sources":[{"id":"s1","url":"https://arstechnica.com/tech-policy/2026/09/chatgpt-using-lawyer-punished-for-citing-fake-testimony-from-made-up-witnesses/","kind":"news_report","access":"read","language":"en","translation_note":"Read in English on 2026-09-15 (11 Sep 2026, Brodkin). Quotes the 9 September order and the 21 August hearing; names the fabricated witnesses.","independence_group":"ars"},{"id":"s2","url":"https://www.straitstimes.com/world/united-states/chatgpt-invented-fake-police-testimony-in-murder-appeal-new-mexico-high-court-says","kind":"wire_report","access":"read","language":"en","translation_note":"Read in English on 2026-09-15 (11 Sep 2026); Straits Times carrying the Reuters wire.","independence_group":"reuters"},{"id":"s3","url":"https://www.abajournal.com/news/article/criminal-defense-attorney-admits-stupidity-over-ai-errors-but-still-receives-sharp-rebuke-from-his-states-high-court","kind":"trade_press","access":"read","language":"en","translation_note":"Read in English on 2026-09-15 (24 Aug 2026). Post-hearing report citing the Santa Fe New Mexican; the fine and contempt were anticipated from the hearing, not yet the written order.","independence_group":"aba-journal"}],"version":1,"ai_roles":["own_use"],"contexts":["work","justice"],"unknowns":["The 9 September 2026 order PDF was not directly inspected (nmcourts.gov search empty; form-based case lookup); the claims about it rest on two chains quoting it.","The underlying docket number is unconfirmed.","The disciplinary-board process and its outcome are pending.","The full set of fabricated or misdescribed citations beyond the named examples is not reported.","The ChatGPT product tier and whether the firm had verification policies are not reported."],"geography":{"basis":"Events: Santa Fe, New Mexico practice and the Doña Ana County murder case. Affected persons: the attorney and his client, both in New Mexico. Court: New Mexico Supreme Court.","court_countries":["US"],"event_countries":["US"],"affected_person_countries":["US"]},"publication":{"basis":"Published under the 2026-09-15 charter as an own-use, work-context adverse consequence with a dated, quantified sanction and an admitted, judicially found AI link. This applies the charter's individual-review rule to the AI-sanction class; the retired ordinary_task_failure gate and a class-level ruling are not required. Aarons is a public officer of the court sanctioned by a public order and is named; his client's name appears in the public caption of the appeal. In a 2026-09-15 second pass the original compound first claim was split: the corroborated contempt/fine/referral stand as c1a, and the Ars-only order elements (removal, bar, struck briefs, reassignment) stand as reported c1b. person_relations recorded in the same pass.","reviewed_on":"2026-09-15"},"ai_involvement":{"basis":"The AI link is admitted and judicially found: Aarons fed a Rev.com transcript and the record into ChatGPT (OpenAI o3), and the court's order addresses the fabricated content in the filed brief; he admitted at the show-cause hearing that he had not verified the output.","status":"supported"},"person_relations":["communicated_with"]},"name":"New Mexico Supreme Court fines attorney Stephen Aarons $5,000 and removes him from a murder appeal over ChatGPT-fabricated testimony","summary":"Santa Fe defense attorney Stephen Aarons used ChatGPT (OpenAI's o3) on a Rev.com transcript and the record to draft briefing in the murder appeal of Oscar Renee Sandoval; the brief contained fabricated witnesses and testimony (including police officers 'Michelle Amarillo' and 'Sanchez' and invented testimony attributed to real people) and misdescribed real precedents. After a 21 August 2026 show-cause hearing at which Aarons admitted he had not verified the output, the New Mexico Supreme Court's written order of 9 September 2026 held him in direct contempt, fined him $5,000 payable to the State Bar Client Protection Fund within 30 days, removed him from the case, barred him from appearing before the court pending a disciplinary-board referral, struck the filed briefs and reassigned the appeal to the public defender.","incidentDate":"2025-08-01","incidentEndDate":"2026-09-09","incidentKind":"bounded_series","incidentDatePrecision":"range","exposurePattern":"single_interaction","reportedDate":"2026-08-24","aiSystem":"ChatGPT (OpenAI o3)","aiProduct":"ChatGPT","aiCompany":"OpenAI","severity":"medium","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["professional_harm","legal_harm","financial_loss"],"harmOutcomeSummary":"A dated, quantified professional sanction for the attorney who relied on ChatGPT's output: $5,000 fine, direct-contempt finding, removal from the case, a bar on appearing before the court pending a disciplinary-board referral, and struck briefs; his client's appeal is delayed as a result.","frameworkFacets":[],"causationStatus":"established","participantUsersAffectedMin":1,"otherPeopleHarmedMin":1,"affectedCountStatus":"exact","affectedCountEvidence":"Two documented harmed people. The user: Aarons, sanctioned (court order reported by Ars Technica and Reuters). Another person's use consequence: his client Sandoval, whose appeal is delayed by the struck briefs and the substitution of counsel (Ars quotes the court's 'your client is the one suffering'). No further harmed persons are reported; the victim's family is not counted (no harm to them is reported).","victimAgeRange":"adult","jurisdiction":"US","platformType":"assistant","outcomeType":"regulatory_action","outcomeStatus":"resolved","primarySourceUrl":"https://arstechnica.com/tech-policy/2026/09/chatgpt-using-lawyer-punished-for-citing-fake-testimony-from-made-up-witnesses/","primarySourceLabel":"Ars Technica (11 Sep 2026) — quotes the 9 Sep order and the 21 Aug show-cause hearing","firstPublishedAt":"2026-09-15T06:16:05.506355+00:00","updatedAt":"2026-09-30T01:17:22.660417+00:00","scopeVersion":"facts-v3","tags":["work","justice","legal-profession","sanction","fabricated-citations","chatgpt","own-use","new-mexico"]}]}