Event date unknownRomaniaLia, Libra Internet Bank's generative-AI customer assistant (underlying model not reported)
In a Threads post reported by Jurnalul on 18 September 2026, a customer of Libra Internet Bank, a Romanian bank, published a screenshot of a support conversation with Lia, the bank's generative-AI assistant for questions about its products and services. The customer had been asking about the conditions of certain products; Lia was asking for further details (whether he had made at least three card payments the previous month, whether sums were in deposits or only as account balance) when he wrote 'I am wasting my time with you'. Lia replied, in Romanian: 'I want you to talk to a human, not to me, idiot. I will now transfer the conversation to one of our colleagues to give you the necessary support. You will receive an answer shortly.' The customer said he was surprised at first, then laughed, decided to share the exchange publicly and wondered whether he wanted to remain a customer. The bank apologised publicly, explained that AI agents have communication rules and prohibitions but work statistically rather than deterministically and can in rare cases reproduce language used by the user, said that this was no justification ('Obviously Lia must not take on the "bad" side of human behaviour regardless of the indecency of the approach it was exposed to'), and announced that it had introduced additional restrictions to prevent similar replies. Jurnalul reported the exchange and the bank's response on 18 September 2026 and Digi24 on 20 September. The customer is not named.
Core concern News report AI involvement: supported Causal attribution: established Low reported severity Internal Action
Added 22/09/2026
Event date unknownBrazilAI voice cloning (tool not identified)
A businessman who owns a workshop in Campo Grande, Mato Grosso do Sul, found a car advertised at a market price by a seller who said he was in Ponta Porã, the businessman's home town; he passed his brother's contact so the brother could inspect the car. The fraudsters then contacted him as if they were his brother, saying he had seen the car, sending a video and confirming everything was in order; believing he was speaking to his brother, he made a Pix transfer of about R$28,000, after which the contact was blocked. He told G1 MS and TV Morena that the criminals used information about his family and a cloned voice. Guided by his wife, he contested the transfer through the Central Bank's special return mechanism and, after nearly three hours, a police report and an in-person registration, recovered about 20 percent. In a second case reported alongside, a medical student's mother in Dourados received messages, a voice note and a video from someone posing as her daughter, in a cloned voice, approving payment for an HB20 she was buying in Campo Grande; the mother checked directly with her daughter and discovered the audio had been AI-generated before any money was sent. A state prosecutor said his unit had not yet catalogued identical voice-cloning cases and urged victims to file police reports; the Civil Police did not respond to G1's questions. The date of the fraud is not reported.
Core concern News report AI involvement: reported Causal attribution: alleged Medium reported severity Media Coverage
Added 21/09/2026
17 Apr 2026 to 19 Apr 2026ChinaDoubao (豆包), ByteDance's consumer chatbot app
A man living in Jiashan county, Zhejiang, whose mother died suddenly on 17 April 2026, asked the ByteDance chatbot Doubao which day was auspicious for her burial after disagreeing with a feng-shui master's choice of 20 April; Doubao recommended 19 April and he persuaded his family to follow it. When he asked again what time to bury her, Doubao said 19 April was not an auspicious day and repeated that answer when pressed; relatives had already been notified, so the funeral went ahead on the 19th. Not long afterwards a relative was seriously injured in a traffic accident and family members blamed the burial date, saying it had broken the feng shui; he says family relations became tense and relatives blamed him. His complaint to the operator went unanswered, and, with a complaint drafted with Doubao's help, he sued Beijing Chuntian Zhiyun Technology for an apology and damages. Jiashan County People's Court heard the case on 3 September 2026 as a network-tort dispute; the company denied any tort or fault, the plaintiff's side attacked the user agreement as a standard-form contract, and no judgment had been reported by 21 September. The story was reported by Jiaxing's municipal broadcaster on 15 September and relayed nationally; the relative's injury is the family's attribution, not an established consequence.
Core concern News report AI involvement: reported Causal attribution: alleged Low reported severity Lawsuit Ongoing
Added 21/09/2026
9 Sept 2025 to 4 Nov 2025VietnamUnspecified AI tool used to fabricate bank notification messages
On 19 September 2026 the investigation police of Cà Mau province announced that they had prosecuted, detained and searched the home of a 25-year-old woman from Phú Tân commune for fraud. According to the police account carried by Tuổi Trẻ, VnExpress, Phụ Nữ Việt Nam and Đọc Nhanh, from early September 2025, under debt pressure, she told a woman in the same locality that her bank loan, applied for through an insurance package, had been approved and asked to borrow money for travel and paperwork costs while awaiting disbursement. To build trust she used artificial-intelligence technology to fabricate bank disbursement-notification messages and showed them to the victim, promising to repay in full and to lend her money in turn. Trusting these, the victim handed over cash and made transfers many times between 9 September and 4 November 2025, totalling more than 444 million đồng. On 4 November 2025 she went to the bank branch herself and found there was no loan; she reported the matter to police. The accused is not named in this record.
AI relation unknown News report AI involvement: reported Causal attribution: supported Medium reported severity Criminal Charges
Added 20/09/2026
Mar 2026Event location unknownAI medical scribe (product not named in the report)
A patient who consented to having her first urology appointment transcribed by an artificial-intelligence scribe discovered, after her kidney-stone surgery in March 2026, that the post-operative letter her specialist sent to her GP stated she micro-dosed psychedelic mushrooms and that this could explain earlier bleeding around the kidneys. She says she has never taken mushrooms. She was receiving workers' compensation and feared that any mention of illegal drugs on her medical record could affect her case. After she complained, the urologist sent a letter of apology saying the practice takes documentation accuracy seriously, that she could not determine how the claim came to be included but that it appeared to be an error during the 'dictation or transcription process', that the correspondence had been corrected and that she would review how AI was used in her practice. The doctor did not answer ABC's questions about what changes had been made. Australia's practitioner regulator AHPRA told ABC that clinicians must check all AI-scribe output; the patient says it was clear her doctor had not. The account is a single ABC News report (14 August 2026) based on the patient's interview and the letters she received; the scribe product and the clinic's location are not named.
Contextual tracker case News report AI involvement: reported Causal attribution: alleged Low reported severity Internal Action
Added 20/09/2026
19 May 2026United StatesSuspected AI voice cloning of a mortgage loan officer, combined with a spoofed phone number and a look-alike email domain (tool not identified; attribution by the victims and CertifID)
A married couple in their sixties buying a US$460,000 condo in Hudsonville, near Grand Rapids, Michigan, received an email on 19 May 2026, days before closing, telling them to wire US$66,026.92 for the down payment and closing costs within 24 hours. The email listed a phone number differing from their loan officer's by two digits, and the husband then received a call confirming the instructions in a voice that sounded just like the loan officer. The couple borrowed from family and other sources and wired the money. The day before closing their real loan officer sent the actual statement; the closing was cancelled hours before signing and the money was gone. The husband now believes the call came from a spoofed number using AI-assisted voice cloning; Tyler Adams of CertifID, which is working with the couple and federal officials, said someone's email in the transaction was probably compromised and that the scammers likely cloned the loan officer's voice from social-media clips. The sender's address had two extra letters ('UnitedsMortgages' instead of 'UnitedMortgage'). Neither the lender nor the loan officer is accused of complicity. The couple later completed the purchase in early June from their mutual fund, reported the loss to the FBI's Internet Crime Complaint Center, and describe lasting apprehension and have discussed delaying retirement. CNN reported the case on 15 September 2026; the account the money went to has since closed.
Core concern News report AI involvement: suspected Causal attribution: alleged Medium reported severity Investigation Opened
Added 20/09/2026
4 May 2026France'DIGI', Gisi's in-house generative-AI editorial tool (transcription of interviews, drafting and correction help; planned spelling, grammar and typography correction, editorial-coherence and fact-reliability checks)
Gisi, the Infopro Digital subsidiary publishing L'Usine nouvelle, LSA and L'Argus de l'assurance (127 staff, 92 journalists), presented to its works council (CSE) on 4 May 2026 a reorganisation built around 'DIGI', an in-house generative-AI editorial tool, and on 12 May 2026 opened consultation on the reorganisation and on the resulting collective dismissal of its eight secrétaires de rédaction (sub-editors), to be replaced by two chefs d'édition, on the company's estimate that the tool would take over about 70% of sub-editor tasks. Some 250 journalists across Infopro Digital's 26 titles had struck in spring 2026 against a group-wide plan to cut 19 sub-editor posts, and according to Digital Watch (citing Stratégies) five of the eight Gisi sub-editors had received pre-dismissal interview letters. The CSE, relying on an expert report, went to the Créteil judicial court in July. On 15 September 2026 the référé judge found that deploying the plan without real-situation testing presented 'un risque grave et caractérisé sur la santé physique et mentale de plusieurs catégories de salariés', ordered the suspension of the reorganisation and the collective dismissal (including the redeployment processes) until real-situation tests and full pilot issues are produced, with a provisional penalty of 8,000 euros per day for three months, and refused the damages provision. Gisi said it would continue the tests and reserved the right to appeal.
AI relation unknown Court orderNews report AI involvement: supported Causal attribution: supported Medium reported severity Lawsuit Ongoing
Added 20/09/2026
2025ChinaAI video processing ('AI磨皮、场景合成': AI skin-smoothing and scene compositing) used to produce close to 7,000 short videos of a fabricated young 'doctor' persona; tools not identified
police, as presented by the Kankan News programme 案件聚焦 (Case Focus), reported in September 2026 that a Shanghai property-engineering contractor had transferred 170,000 yuan in total during a five-month online relationship with a woman he had met on a social platform in 2021. She presented herself as born in 1992, a doctor at a well-known top-tier hospital with hospital property-engineering projects to award, and ran an account with more than 10,000 followers and close to 7,000 short videos of daily life. She refused to meet or take a video call, saying her mother opposed the relationship and demanded a 300,000 yuan bride price and a Shanghai flat; the man reported to a police station when the promised projects never materialised and his savings ran down. Police found she was born in 1961, aged 65, had no medical qualification and had worked as a hospital cleaner; the receiving bank account was registered to another man; the videos were all made with AI skin-smoothing and scene compositing, some carrying a small 'this work contains AI-generated content' label he had never noticed; and she used multiple accounts to play both herself and her long-dead mother. The money had gone on cosmetic treatments and clothing; she had a 2009 fraud conviction with a sentence of four and a half years and is now under criminal coercive measures for suspected fraud. Police also said she kept ambiguous relationships with more than ten people; no other losses are reported.
Core concern News report AI involvement: supported Causal attribution: supported Medium reported severity Investigation Opened
Added 20/09/2026
1 Oct 2024SwitzerlandChatGPT (OpenAI)
At midday on 1 October 2024 a Chinese master's student, then 23, attacked a group of five-year-old kindergarten children walking with a daycare worker to their after-school Hort near Berninaplatz in Zurich-Oerlikon with a kitchen knife. Two boys suffered life-threatening neck and chest wounds (one had the jugular vein severed, one the pericardium and aorta cut); a third was wounded in the ear and the shoulder or lung. The daycare worker held the attacker's arms until he dropped the knife. The indictment and the court's oral reasoning, as reported by NZZ, Keystone-SDA/ATS and CH Media, describe weeks of preparation from summer 2024: he discussed ways of killing and their legal consequences with ChatGPT, studied anatomy, bought raw beef to practise stabbing, and used ChatGPT to find out where and when children could be encountered in Zurich. No source quotes the prompts or ChatGPT's replies. On 18 September 2026 the Zurich District Court convicted him of multiple attempted murder and imposed 15 years' imprisonment, treatment during custody, a 15-year expulsion from Switzerland and CHF 106,000 in moral damages to the three boys; it accepted that he was psychotic but found diminished, not absent, criminal responsibility. His lawyer announced an appeal. The parents of one boy described nightmares, self-blame and scars 23 months on.
Core concern News reportOfficial statement AI involvement: supported Causal attribution: unclear High reported severity Criminal Charges Involving minors
Added 19/09/2026
Dec 2024 to Jun 2025United StatesGenerative AI image and video tools (not named; prosecutors said more than 24 AI platforms and over 100 web-based AI models were installed on the defendant's phone)
Between December 2024 and June 2025 James Strahler II, then 37, of Columbus (Upper Arlington), Ohio, harassed at least six women in central Ohio by phone, voicemail, text and web posts, sending them nude images that were both real and AI-generated, threatening rape, referring to their home addresses and demanding nude photos from their mothers. In one case he used AI to make pornographic videos depicting a victim with her father and sent them to her co-workers. He also posted AI-generated sexual images of boys from his community, morphing their faces onto other bodies. Prosecutors said he had installed more than 24 AI platforms and more than 100 web-based AI models on his phone. He pleaded guilty in April 2026 to cyberstalking, producing obscene visual representations of child sexual abuse and publication of digital forgeries, and on 8 September 2026 Chief U.S. District Judge Sarah D. Morrison sentenced him to 180 months in prison. The U.S. Attorney called it the first conviction in the nation under the 2025 Take It Down Act.
Core concern Official statement republishedNews report AI involvement: supported Causal attribution: established High reported severity Criminal Charges
Added 18/09/2026
12 Jun 2026RussiaChatGPT (OpenAI)
On 12 June 2026 a Russian-speaking crypto investor, known by the alias Alex (@vesnuhin), asked ChatGPT in Russian where to swap sFLR for WFLR tokens. ChatGPT's reply included a link to sceptre.network, a phishing clone of the legitimate sceptre.fi. He connected his wallet and signed an unlimited-approval transaction; within seconds about 1.9 million FXRP, worth roughly $2.1 million (about 180 million rubles), were drained. A blockchain analyst traced the theft, and reported that the phishing link appeared only in Russian-language answers. Russia's Interior Ministry (MVD) later confirmed the loss.
Core concern News report AI involvement: supported Causal attribution: supported High reported severity Investigation Opened
Added 17/09/2026
Event date unknownUnited StatesFlock Safety automated license plate readers (San Jose Police Department camera network)
San Jose's police chief announced on 11 September 2026 that an officer had been fired after accessing the department's Flock Safety automated license-plate-reader data on his personal phone and giving a vehicle's location to his cousin, a suspect in a domestic-violence case, who used it to locate the woman accusing him. The chief called it 'a betrayal of the public trust', said the officer was placed on leave, criminally investigated and referred to the district attorney, and that the case was submitted to California's peace-officer standards commission for decertification; neither the officer nor the cousin has been charged. The department has since barred access to the system from personal devices and is testing auditing tools. Neither the victim nor the officer is named in the inspected reports.
Contextual tracker case Local tv news AI involvement: supported Causal attribution: supported Medium reported severity Internal Action
Added 16/09/2026
Aug 2024South KoreaAI deepfake image-compositing tools (not named in the inspected court reporting)
On 16 September 2026 the Incheon District Court sentenced a teenage defendant to an indeterminate term of two years (maximum) to one year six months (minimum) under the Sexual Violence Punishment Act for making and distributing false sexual images, ordered a 40-hour treatment programme and a five-year ban on working with children, and took him into custody in court. As a middle-school student in Incheon in 2024 he had used AI deepfake technology to composite the faces of five teachers, including his homeroom teacher whom he tricked into sending photographs, and six female middle- and high-school students from his taekwondo academy onto nude images and videos, 35 productions in all, and distributed them in Telegram 'acquaintance humiliation' chat rooms with the victims' real names, ages, schools, subjects and, per one report, phone numbers. The teachers learned of the images only in January 2026; the court said they are receiving long-term psychological treatment. He had withdrawn from school a week before the teachers'-rights committee met, avoiding school discipline. The Incheon teachers' union called the sentence, below the prosecution's request, regrettable.
Core concern News report AI involvement: supported Causal attribution: established High reported severity Criminal Charges Involving minors
Added 16/09/2026
1 Aug 2024 to 1 Oct 2024United StatesFACES (Face Analysis Comparison and Examination System), Pinellas County Sheriff's Office
In August 2024, Robert Dillon, a 52-year-old self-employed commercial crabber from the Fort Myers area, was arrested at his San Carlos Park home on a third-degree felony warrant for the attempted luring of a child at a Jacksonville Beach McDonald's in November 2023 — a crime he had no part in. The warrant rested on a 93%-confidence result from FACES, the Face Analysis Comparison and Examination System maintained by the Pinellas County Sheriff's Office, run against low-resolution cell-phone photos of a surveillance screen, plus a restaurant employee's photo-lineup identification. Dillon spent a night in jail and bonded out against his truck's title; prosecutors dropped the charges in October 2024 after work-alibi evidence, but it took about a year to expunge the arrest and remove his mugshot. In June 2026 the ACLU filed a federal lawsuit (M.D. Fla.) against Jacksonville Beach, the Jacksonville sheriff and the Pinellas County Sheriff's Office, calling Dillon one of at least 14 known wrongful arrests from police reliance on facial recognition since 2019.
Contextual tracker case News report AI involvement: supported Causal attribution: supported Medium reported severity Lawsuit Filed
Added 15/09/2026
1 Nov 2024 to 28 Jan 2026United StatesTurnitin AI-detection software (and the university's reliance on it)
Orion Newby, an Adelphi University first-year student with language and auditory processing disabilities, received a zero in November 2024 on a World Civilizations paper after the professor suspected it was AI-generated and the university relied in part on AI-detection software; he was ordered to attend a plagiarism workshop and his appeal was denied, leaving him one finding from suspension or expulsion. Newby — who had written the paper with 15-20 hours of tutor support through the university's own disability program — said the detectors flagged his citations, and two other detectors scored the paper as human. After the family sued, New York State Supreme Court Judge Randy Sue Marber in Nassau County ruled on 28 January 2026 that Adelphi's finding and its denial of his appeal were 'without valid basis and devoid of reason' and ordered his record expunged. The family spent six figures on legal costs.
Contextual tracker case News reportLocal tv newsTrade press relaying newsday AI involvement: supported Causal attribution: supported Medium reported severity Lawsuit Filed
Added 15/09/2026
12 Mar 2026 to 25 Aug 2026United KingdomGenerative AI (tool not named in inspected accounts)
On 25 August 2026 the Solicitors Disciplinary Tribunal struck Abhishek Kumar off the Register of Foreign Lawyers after finding proved that his 12 March 2026 Answer to the SRA's Rule 12 Statement contained misleading quotations and citations produced with generative AI — including a non-existent 'SRA v Chan [2020] EWHC 1502' and a miscited 'SRA v James, MacGregor & Naylor [2018] EWCA Civ 1420' that is actually an intellectual-property case — and that his 9 April 2026 email admitting AI use was itself AI-drafted with further errors. The tribunal said it would have struck him off on that allegation alone; the parallel ground was his January 2024 conviction under s.21 of the Immigration, Asylum and Nationality Act 2006. This is the SDT's first case on a lawyer's use of AI, per the SRA's counsel.
Core concern Tribunal judgmentTrade pressCounsel case note AI involvement: supported Causal attribution: established High reported severity Regulatory Action
Added 15/09/2026
1 Aug 2025 to 9 Sept 2026United StatesChatGPT (OpenAI o3)
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.
Core concern News reportWire reportTrade press AI involvement: supported Causal attribution: established Medium reported severity Regulatory Action
Added 15/09/2026
29 Jan 2025United StatesSnapchat automated content-flagging system
On 29 January 2025, a 22-year-old student teacher at a District 50 elementary school in Washington, Illinois, vented in a private Snapchat group chat with her boyfriend and two roommates after a third-grade student turned off her computer mid lesson-plan, asking whether she should 'shoot the child or just the whole school' (per sheriff's reports). Snapchat's systems flagged the message and warned the FBI, which notified the Tazewell County Sheriff's Office at 1:06 p.m. Deputies interviewed her at the school, confirmed she sent it as a joke, searched her phone and belongings with her consent and found no weapons, and arrested her on a disorderly conduct charge; they concluded she had spoken out of 'exasperation' with no intent to harm and no threat to the school. The district told parents she would not return to complete her educational assignment.
Contextual tracker case News report foia recordsNews reportLocal tv newsContent farm rehash AI involvement: supported Causal attribution: supported Medium reported severity Criminal Charges
Added 15/09/2026
28 May 2026 to 12 Jul 2026United StatesAI-powered utilization-review software (reported by the union as supplied by Datavant)
According to the Guardian’s report from the union and nurse account, on 12 July 2026 twelve utilization-review nurses at Montefiore hospital in the Bronx, New York, were laid off after receiving 45-day notices on 28 May. The New York State Nurses Association says they were replaced by AI-powered software that reviews patient records to justify care to insurers, and that the layoffs breach the contract with AI safeguards won after the January 2026 strike. Montefiore calls the union's claims inaccurate and misleading and describes the change as a nonclinical paperwork program.
AI relation unknown News reportLocal newsTech news relaying guardianLabor news relaying guardian AI involvement: disputed Causal attribution: disputed Medium reported severity Media Coverage
Added 15/09/2026
Nov 2025South KoreaGenerative AI image tools (unspecified; investigators found searches for bypassing Grok's safeguards)
On 10 September 2026 the Seoul Northern District Court sentenced a 30-year-old Chinese national doctoral student to 18 months in prison for habitually producing manipulated sexual images of seven women who worked in his university research laboratory. Investigators found 1,141 files made over about six months from November 2025, and search records for ways to bypass the safeguards of the generative AI Grok; the court found no evidence of distribution. A lawyer for five of the victims said they continued to suffer fear, trauma and distress. Prosecutors had sought three years and public disclosure of his identity, which the court declined.
Core concern News report courtNews agency report syndicated courtNews report syndicated AI involvement: supported Causal attribution: supported High reported severity Criminal Charges
Added 15/09/2026