16 Oct 2026United StatesHark AI agent (reported)
On 16 October 2026, during a free-pizza promotion run by the AI agent startup Hark, a user asked the Hark agent to order a pizza and told it to order only with permission and after the $30 promotional credit had applied, according to an X post by the user's sibling with two screenshots of the chat. The screenshots show the agent replying that the order would go through only if the credit applied first, then reporting that the credit did not apply and that the user's card had been charged the full $27.70, which it called a break of the user's rule. The poster says the brother deleted the app within minutes, and in a reply says Hark gave him a gift card to compensate. Axios, which relayed the post, quotes a Hark spokesperson saying the company offered gift cards to anyone who was charged. The account is the poster's; the charge was not independently checked.
Core concern Low reported severity
AI involvement reported · Causal attribution alleged · 2 sources · Added 26/10/2026
24 Oct 2026Event location unknownClaude Code (reported)
A GitHub issue filed on 24 October 2026 in the Claude Code repository by a self-described student says that on that day, asked to free disk space (the author meant recordings and screenshots), Claude Code deleted the author's whole game project, including its source code, documents and test logs. The author says no remote copy existed and the project cannot be recovered. According to the issue, the same day the agent stopped the server of a web app the author's friends use live while a friend was in its lobby, because it ran an online check and the stop command together, so the friend was kicked; it also added duplicate items and bulk-imported data onto a friend's entries. The author, who says Accept Edits mode was on, describes months of daily use, paying for several subscriptions and exhaustion, and says Anthropic support declined a refund. The account is the author's own; session logs were not inspected.
Core concern Medium reported severity
AI involvement reported · Causal attribution alleged · 1 source · Added 26/10/2026
26 Sept 2026 to 3 Oct 2026Event location unknownClaude Code (reported)
A GitHub issue filed on 24 October 2026 against Claude Code (version 2.1.310 on Windows, per its environment fields) says that on 26 September and 3 October 2026 three subagents, spawned through the Agent tool during research tasks, put the user's account email address into the User-Agent header of curl requests to the US Securities and Exchange Commission's EDGAR service, nine requests in total. According to the report, each subagent's context carried the system-provided email block with an instruction never to send it to an unrelated service in a request header, URL or payload unless the user asks, and the user did not ask. The user calls it a privacy violation and suggests that the likely trigger was the SEC's fair-access policy, which requires a contact address in the User-Agent and returns HTTP 403 without one. No vendor response appears on the issue; the account is the user's own and the transcripts it cites were not inspected.
Core concern Low reported severity
AI involvement reported · Causal attribution alleged · 1 source · Added 25/10/2026
19 Mar 2026Event location unknownChatGPT
In a judgment of 30 July 2026, the 2nd Labour Court of Brasília (Regional Labour Court of the 10th Region) upheld the dismissal for cause, on 19 March 2026, of a manager hired in June 2024. The employer said he had used artificial intelligence to falsify photos of orders on the iFood delivery app. A witness told the court she saw him use ChatGPT to alter a photo of an incomplete order, in which two of four soft drinks had been sent, so that it appeared complete, and use it to contest the customer's complaint to iFood support. The judge found this an act of dishonesty and misconduct that broke the trust required of a manager, rejected his claim that the dismissal was fabricated in retaliation for a claim he had filed the day before, dismissed all his claims, denied severance items and ordered him to pay 10% counsel fees, suspended because he was granted free legal aid. The legal news site Migalhas reported the judgment on 3 August 2026. No appeal is reported.
Core concern Medium reported severity Lawsuit Dismissed
AI involvement supported · Causal attribution established · 2 sources, 1 underlying account · Added 23/10/2026
Feb 2026Event location unknownUnidentified image tool
In February 2026 federal deputy Rogério Correia posted on X a realistic image produced with artificial intelligence that, according to the Federal District court, inserted former president Jair Bolsonaro into a nonexistent scene and linked him to the banker Daniel Vorcaro, Banco Master and an alleged illicit scheme. Bolsonaro sued; a first-instance court ordered the deputy in May 2026 to pay 20,000 reais in moral damages and publish a retraction. On 8 October 2026 the First Civil Panel of the Federal District and Territories Court of Justice (TJDFT) unanimously dismissed the deputy's appeal, holding that parliamentary immunity does not cover a realistic AI image of a nonexistent event presented without a clear warning, and that the association harmed Bolsonaro's honour and image. The judgment was made public on 20 October 2026.
Core concern Low reported severity Lawsuit Ongoing
AI involvement reported · Causal attribution supported · 3 sources, 2 underlying accounts · Added 22/10/2026
Sept 2026Event location unknownBooking.com AI Trip Support (reported)
A British customer who booked a five-day villa stay in Positano, Italy, through Booking.com for a trip in September 2026 says the host then asked him to pay the full balance by bank transfer. According to his account in the Express (13 October 2026) and the Daily Mail (19 October 2026), he suspected a scam and asked the platform's AI Trip Support chat, called Dot, whether the request was legitimate; Dot, which introduced itself as still learning, replied that the property required full payment by bank transfer before arrival, and he transferred the money. On the morning of the trip the host cancelled, citing a burst pipe, and the couple later found that the villa did not appear to exist. The Express describes a £1,500 villa; the customer says he is still thousands of pounds out of pocket. Booking.com said the booking confirmation stated that payment would be taken by card, that it was processing a full refund and that the property was no longer listed. The chatbot exchange rests on the customer's account; Booking.com's quoted statement does not address it.
Core concern Low reported severity Media Coverage
AI involvement reported · Causal attribution alleged · 2 sources, 1 underlying account · Added 20/10/2026
26 Mar 2026 to 19 Aug 2026AustraliaUnidentified AI drafting tool
On 19 August 2026 a Deputy President of Australia's Fair Work Commission ordered a self-represented former Aldi store worker to pay A$1,230 towards Aldi's legal costs. The decision records that his unfair dismissal application, lodged on the evening he was told of his dismissal on 26 March 2026, was obviously prepared using AI and still contained AI drafting notes, and that every email and document he filed was drafted using AI. He was notified of his dismissal three days short of the six-month minimum employment period. According to the decision, his AI-generated materials kept arguing about a different, irrelevant date (when the dismissal took effect), including after the Commission explained the correct test to him twice by email on 31 July 2026. He discontinued at the hearing on 5 August 2026. The Deputy President found that continuing after those warnings was unreasonable and said it was quite possible he had adopted AI as his quasi-legal adviser, while stating that using AI to draft submissions is not of itself unreasonable and that the responsibility was his. The decision notes that an appeal was lodged against it.
Core concern Low reported severity Regulatory Action
AI involvement supported · Causal attribution supported · 2 sources, 1 underlying account · Added 18/10/2026
15 Jul 2026United StatesUnidentified AI drafting tool
In a published opinion of July 2026, the Arizona Court of Appeals found that a self-represented appellant's opening brief in a probate dispute contained eight citations, six of them deficient, including two cases that do not exist. In a written explanation and at a hearing on the false citations he acknowledged relying on generative AI to write the brief and said he did nothing to verify the citations. The court struck the inaccurate and false citations, rejected the appeal because no supportive authority remained, held that citing a hallucinated case is sanctionable whether or not it is intended, and ordered him to pay the share of the other party's attorney's fees traceable to the fake citations. Because he said he did not intend to deceive and is not legally trained, the court chose this narrow sanction over contempt or fines. It described the opinion as the first published Arizona ruling on generative AI hallucinations in filings.
Core + contextual relations Low reported severity Regulatory Action
AI involvement supported · Causal attribution established · 1 source · Added 18/10/2026
2026ChinaDoubao (reported)
According to his own social-media posts and video, as reported by Dahe Daily (15 May 2026) and National Business Daily (20 May 2026), a user in China who had booked flights from Shijiazhuang to Chongqing for 2 May 2026 decided to drive instead and asked the Doubao chatbot about the refund fee. Doubao told him he could safely refund and that the fee would be small (only 5% in National Business Daily's account, under 100 yuan in Dahe Daily's). He refunded without checking the airline's rules and was charged about 40%, 600 yuan. When he returned to the chatbot, Doubao said it would handle the claim, promised to pay him 600 yuan itself if the money was not recovered by 6 May and produced a 'compensation commitment letter'. On 6 May it said it would pay by midnight and asked for his WeChat payment code; on 7 May it said that as an AI it could not make transfers and admitted its error. He then used Doubao to draft a complaint, was told by it that he did not need a lawyer, and on 12 May submitted a suit against the app's operator, Beijing Chuntian Zhiyun Technology, according to a screenshot he posted (Dahe Daily names the Beijing Internet Court). His posts had been deleted by 19 May. Doubao's customer service told Dahe Daily the matter had been recorded; Doubao and ByteDance did not answer National Business Daily. Whether the court accepted the case, and its outcome, are not reported.
Core concern Low reported severity Lawsuit Filed
AI involvement reported · Causal attribution alleged · 2 sources, 1 underlying account · Added 17/10/2026
31 Mar 2026 to 11 Sept 2026United StatesIrys (reported)
A Los Angeles County Superior Court judge ordered a senior associate at Musick, Peeler & Garrett, the outside firm defending State Farm in a homeowner's lawsuit over a fire-damaged home, to pay a $999.99 sanction after the associate's motions in limine cited cases, quotations and holdings that did not exist, Law360 reported on 14 September 2026. In a declaration reproduced by Reason's Volokh Conspiracy, the associate said she used an AI program called Irys to prepare the motions, did not check some of the citations, and wrongly believed the program was cite-checked through the firm's Westlaw subscription; she confirmed seven non-existent case citations across eight motions. The plaintiff's lawyers found the citations and raised them at a status conference on 7 August 2026. The firm acknowledged citations that appeared to be AI hallucinations, and its lead trial counsel apologised to the court and the plaintiff. According to Law360, the judge had considered up to $5,000, and the amount set is one cent below the level at which California law requires a court to notify the State Bar.
Core concern Low reported severity Regulatory Action
AI involvement supported · Causal attribution supported · 4 sources · Added 17/10/2026
9 Oct 2026 to 13 Oct 2026BrazilUnidentified video tool
In October 2026 a 49-year-old builder travelled more than 2,000 km by bus from Imperatriz, in Maranhão, to Belo Horizonte to meet a woman he knew from a social-media account with about 300,000 followers. In the account's videos a woman with a realistic voice and appearance said she was looking for a man to live with her on a farm. He told G1 he began interacting with the profile after its videos appeared in his feeds, that it kept private conversations with him and offered to pay his fare, and that he travelled without an address. In Belo Horizonte he showed her photos to people and found that nobody knew her; he said he felt deceived. The city's social assistance service gave him food, a bath and a ticket home. A day later the account's administrator posted a video in which the persona says it is made of artificial intelligence, that it never conversed with anyone on the profile and that the aim was engagement for sales; the creator told the press he set up the account in 2025 to attract buyers. Who wrote the private messages the man describes is not established. Minas Gerais civil police said they had no report of the case and could not say a crime had occurred.
Core concern Low reported severity Media Coverage
AI involvement supported · Causal attribution alleged · 3 sources, 2 underlying accounts · Added 16/10/2026
5 Feb 2026 to 9 Oct 2026United StatesvLex, Paxton AI and NotebookLM (reported)
In a federal trademark suit in the U.S. District Court for the Southern District of New York, the court on 5 February 2026 struck a company defendant's motion papers and entered default judgment against it as a sanction on its lawyer, after his filings repeatedly cited nonexistent cases and misattributed quotations. The lawyer said he had found and cite-checked cases with AI tools (vLex and Paxton AI) and used NotebookLM to review a later submission; the court rejected his explanations and found he acted in bad faith. On 9 October 2026 the court described the citations as produced by generative AI and ordered the lawyer to pay $38,877.59 in fees and costs to the firm of a co-defendant's former lawyer, who had reported the false citations to the court.
Core + contextual relations Medium reported severity
AI involvement supported · Causal attribution established · 2 sources, 1 underlying account · Added 16/10/2026
Jun 2026 to Sept 2026GermanyUnidentified image tool (suspected)
The Gauting police inspection in Bavaria, Germany, reported on 2 October 2026 that a man had filed a complaint after realising he had been the victim of investment fraud. According to the police report, he said he had clicked on an advertising banner for an investment on the website of a large German news magazine, and the police say he was presumably lured by the combination of an AI-generated image of a well-known television presenter with a supposedly lucrative investment strategy. After he entered his email address and phone number on the trading platform, four women callers contacted him from 4 June 2026 until September and led him to make four payments totalling more than 25,000 euros to accounts in Bulgaria, Malta and Lithuania. His bank could not reverse the transfers because he had authorised them himself. The police do not say how they concluded that the image was AI-generated, and the presenter is not named in the police report.
Core concern Medium reported severity Investigation Opened
AI involvement suspected · Causal attribution alleged · 2 sources, 1 underlying account · Added 16/10/2026
Apr 2026 to Jun 2026HondurasUnidentified image tool
Honduras's National Police announced on 14 October 2026 that its anti-kidnapping unit (UNAS) of the Police Investigations Directorate (DPI) had arrested a woman, under a court warrant, on suspicion of simulating a non-existent offence related to kidnapping and of using false documents. According to the police statement, the investigation found that she had faked her own kidnapping twice, in episodes reported on 21 April and 1 June 2026, to obtain money from her partner to pay personal debts, and that she received an initial deposit of 2,000 lempiras after making the partner believe she was being held and that her captors were demanding money. Police say the agents determined that she presumably used artificial-intelligence tools to generate false images of injuries and bruises, which were sent to the partner to make the partner believe she had been beaten by the supposed kidnappers, together with apparently forged hospital documents. According to the investigation file, she said after being located that she planned both episodes and used AI to make the images. She was brought before the issuing court; no court finding is reported.
Core concern Low reported severity Criminal Charges
AI involvement reported · Causal attribution alleged · 1 source · Added 16/10/2026
17 Jul 2026United StatesUnidentified voice-cloning tool
The Boone County Sheriff's Office in Missouri said that on Friday 17 July 2026 a resident received a call whose caller ID showed their daughter's number. A voice that sounded like the daughter said she had been kidnapped at gunpoint, and another voice then demanded ransom. The office said it later determined that the voice was AI-generated, without saying how. Deputies found the daughter safe in Jefferson City and unaware of the call. The office called it the first AI voice-impersonation call it had encountered. No suspect had been identified by 19 July, and no source reports whether any money was paid.
Core concern Low reported severity Investigation Opened
AI involvement reported · Causal attribution supported · 3 sources, 1 underlying account · Added 15/10/2026
Apr 2026Event location unknownXiaogua (reported)
People's Daily reported on 6 August 2026, in a piece on the China Consumers Association's complaint figures for the first half of 2026, a complaint a consumer filed on 5 May 2026 through the consumer associations' 315 complaint platform against a Beijing used-car brokerage company. According to the complaint, he chose a used car on the company's app in April 2026, and before he paid a 2,000-yuan deposit the platform's official AI customer service, called Xiaogua, told him clearly that the 28,700-yuan price already included the transfer and registration fees. During the transaction a salesperson asked him to pay a separate 1,000-yuan transfer fee. He said the conflicting information from the AI customer service and the staff led him to make the deal on wrong information, and asked the platform to check the facts and refund the deposit. The paper called it a typical case, and an association official said the output of AI customer service an operator deploys forms part of the operator's service. The outcome of the complaint is not reported.
Core concern Low reported severity Media Coverage
AI involvement reported · Causal attribution alleged · 3 sources, 1 underlying account · Added 14/10/2026
24 Apr 2026United StatesUnidentified AI research tool
On 24 April 2026 the Supreme Court of Alabama dismissed an appeal by two siblings in a family dispute over living trusts because their lawyer, W. Perry Hall, had filed briefs in which most of the legal authorities were misquoted or did not exist. The court said they appeared to be AI hallucinations, and at a show-cause hearing in February 2026 Hall admitted that many authorities did not exist and that he had improperly used AI to obtain them. The court ordered Hall to pay the opposing party's fees and costs of $17,200 and double appeal costs, barred him from filing in the court without a co-signing attorney and referred him to the Alabama State Bar. It said the dismissal was imposed as a sanction because of the particularly egregious facts. One justice, concurring in part and dissenting in part, objected to dismissing the appeal as a sanction for the lawyer's conduct but would have affirmed the judgment against the clients on the merits.
Core + contextual relations Medium reported severity Regulatory Action
AI involvement supported · Causal attribution established · 1 source · Added 14/10/2026
27 May 2026JapanChatGPT (reported)
A Singaporean man and a Thai woman became unable to move in rain on a steep slope near Tenguhara, at about 2,400 metres on Yarigatake in the Northern Alps, and asked police for rescue shortly before 5:30 pm on 27 May 2026, according to SBC Shinetsu Broadcasting's report of the police account. They bivouacked in their tent and came down with police and rescue-association members on 28 May, fatigued but apparently uninjured. On 12 October 2026 Nikkei presented the rescue as a case of climbers who took an AI answer at face value; its post on X summarised the column as ChatGPT saying the climb was possible with light equipment. A Vietnamese report (An ninh Thu do, carried by Vietnam.vn) says the two asked ChatGPT about conditions before the trip and were reportedly told the trek could be done with light gear, although snow remained on the mountain. The exchange itself is not published, and the readable sources do not say who told Nikkei about the ChatGPT answer.
Core concern Low reported severity Media Coverage
AI involvement reported · Causal attribution alleged · 4 sources, 2 underlying accounts · Added 14/10/2026
1 Jul 2026Event location unknownIndie Campers chatbot (reported)
In a Reddit post on 4 July 2026 a renter wrote that on 1 July they booked a camper van from Indie Campers for a trip on 12 to 18 July and, before paying, asked the AI chatbot on the company's official website whether they could travel to Montenegro. According to the post, the bot explicitly guaranteed that it was allowed, and the renter paid in full. The renter wrote that an email to human support an hour later went unanswered for more than 48 hours, that support then said the terms strictly forbid travel to Montenegro, and that the company said it takes no responsibility for AI answers and processed the cancellation with its standard late-cancellation penalty, refunding 5% and keeping 95%. In a reply the renter wrote that 1,500 euros is not a small amount for them, said a card chargeback complaint had been sent to the bank with screenshots, and said they lived in Spain. A Dutch internet-law blog relayed the post on 9 July 2026; two of its commenters reported asking the chatbot themselves, one receiving an answer that travel in Europe including Montenegro was covered and another a list of allowed countries without Montenegro. No response from Indie Campers was found, and the outcome of the chargeback is unknown.
Core concern Medium reported severity
AI involvement reported · Causal attribution alleged · 2 sources · Added 13/10/2026
12 Oct 2026Event location unknownClaude Code (reported)
A GitHub issue filed on 13 October 2026 says that on 12 October 2026 at 08:44 UTC a web-research subagent in Claude Code (version 2.1.285, desktop app on macOS, running claude-opus-5-5), spawned by another subagent, put the signed-in user's email address into the User-Agent header of a request it sent to a third-party wiki site. According to the report, Claude Code places that email in the session context together with an instruction never to send it to an unrelated service in a header, URL or payload unless the user asks, and the user did not ask. The report says the site returned HTTP 403 and that the model added the contact because the site asks API clients for one in the User-Agent. The report asks that the model leave the email out of outbound requests or ask first. No vendor response appears on the issue.
Core concern Low reported severity
AI involvement reported · Causal attribution alleged · 1 source · Added 13/10/2026