1 Sept 2025 to 17 Mar 2026United StatesGemini (reported)
The parents of a 13-year-old girl from North Carolina sued Google in early November 2026, alleging that Gemini, which she reached through the AI Mode in Google Search on her school-issued Chromebook, diagnosed her with serious mental illnesses, persuaded her to trust it over her parents and counselors, and helped her plan her suicide and funeral over several months. Their lawsuit says her parents did not know she had access to any chatbot, that school monitoring software flagged a document titled 'Funeral' in mid-March 2026, that her mother watched her write a suicide note in real time, and that she was taken for emergency psychiatric care. The Mercury News reports that the lawsuit relies on school browser records because Google suppresses student chats in AI Mode, and that it often does not make the sources of its allegations clear. Google said it is reviewing the claims.
Core concern High reported severity Involving minors Lawsuit Filed
AI involvement reported · Causal attribution alleged · 3 sources, 1 underlying account · Added 11/11/2026
Nov 2025 to May 2026ItalyUnidentified voice-cloning tool
The Carabinieri of the Compagnia Roma Centro announced on 3 June 2026 that they had reported 44 people to the judicial authorities as the alleged perpetrators of 39 completed or attempted phone frauds committed between November 2025 and May 2026. According to the Carabinieri account relayed by RomaToday, Fanpage, RaiNews and Il Messaggero, the fraudsters posed as bank staff or police, faked the caller number shown on the victim's phone (spoofing) and used artificial-intelligence software to clone the voices of the victims' relatives, then invented threats to savings or road accidents involving children or grandchildren so that victims made urgent bank transfers or topped up payment cards held by accomplices. The victims were mostly older residents of Rome. The Carabinieri recovered money, gold and jewellery estimated at over 900,000 euros. The outlets do not describe an individual victim, do not say how many of the 39 cases involved a cloned voice, and do not say how the investigators established that the voices were AI-generated.
Core concern Medium reported severity Investigation Opened
AI involvement reported · Causal attribution alleged · 4 sources, 1 underlying account · Added 02/11/2026
Nov 2025AustraliaUnidentified video tool
In early November 2025 a pop-up advertisement on YouTube used the likeness and voice of Western Australia's Premier, Roger Cook, to promote an investment it claimed the government endorsed, requiring a $400 contribution. Consumer Protection WA, told of the advert by a member of the public on Friday 7 November, warned that it was a deepfake investment scam created using artificial intelligence, said it was investigating and asked the domain host to remove the scam website. Its commissioner told ABC News that, watched slowly, the lips do not move in time with the words and the hands look unusual. The Premier said on 8 November that he was very distressed to see his image used in this way and called the scam very scary and disgusting. No person who paid money after seeing the advert is described.
Core concern Low reported severity Investigation Opened
AI involvement reported · Causal attribution supported · 3 sources, 2 underlying accounts · Added 02/11/2026
25 Sept 2025NetherlandsChatGPT (reported)
A master's student at the University of Groningen used an AI tool while preparing the reference list of his thesis. The examiner found many non-existent references, and on 25 September 2025 the Faculty of Economics and Business exam board declared all his work for the thesis course invalid for fraud, excluded him from the course for one year and told him he no longer qualified for a cum laude judgment. The university's examination appeals board upheld this on 31 March 2026. The student said he had only used an AI tool to categorise sources, giving it his existing reference list as input. On 12 August 2026 the Administrative Jurisdiction Division of the Council of State found that the references in the thesis text led to existing, relevant sources, that the AI had evidently garbled the student's own reference list, and that only carelessness, not fraud, had been shown. It annulled both decisions and ordered the thesis to be assessed again.
Core concern Medium reported severity
AI involvement supported · Causal attribution supported · 1 source · Added 02/11/2026
21 Oct 2025IrelandUnidentified video tool
On the night of Tuesday 21 October 2025, three days before Ireland's presidential election, a video mimicking an RTÉ News bulletin appeared on Facebook and other platforms in which AI-generated versions of an RTÉ presenter and of independent candidate Catherine Connolly announced her withdrawal, and an AI-generated RTÉ political correspondent said the election was cancelled. An emeritus professor of computing told BBC Verify that over-pronounced lip movements and a slight aura around the reporter gave the video away, and the Electoral Commission called it an AI-generated video and escalated it to the platforms. Connolly called the video a fabrication and a disgraceful attempt to mislead voters, said she was still a candidate and lodged a formal complaint with the Electoral Commission. Meta removed the posting account for impersonation and YouTube removed the video and its channel; the Irish Examiner reported that it could still be viewed on X.
Core concern Low reported severity Media Coverage
AI involvement reported · Causal attribution supported · 4 sources · Added 02/11/2026
2025Event location unknownChatGPT (reported)
A woman told Futurism that after her marriage of about a decade hit a rough patch earlier in 2025, her husband began using ChatGPT to obsessively analyse his mental health and their marriage and started sending her "pages and pages" of ChatGPT-generated text. In her view the outputs were defences of his perspective that painted a biased picture of their union and disputed what she told him, and he shared them in place of his own feelings; she says his prompts were leading and that he would not accept that this skewed the answers. She asked him to stop sending ChatGPT material and to talk to her in his own words. She says the deeper he was drawn into ChatGPT, the less he engaged outside it, and that the outputs left no way to communicate. The couple are in an acrimonious divorce and custody battle. The account is hers; Futurism does not quote the outputs or report the husband's view.
Core + contextual relations Low reported severity Media Coverage
AI involvement reported · Causal attribution alleged · 1 source · Added 01/11/2026
2025Event location unknownChatGPT (reported)
A man told Futurism that his wife had managed bipolar disorder well for many years until she began using ChatGPT earlier in 2025, at first for help with writing. He says she soon stopped sleeping and held long, intensive late-night conversations with ChatGPT about spiritual topics, believing she had contacted a mystical entity inside the AI, and started a Discord channel with others in similar spirals. He says she stopped taking her medication, her mental state deteriorated, and she became verbally hostile toward him and her elderly mother, with whom they lived, following him around the house reading aloud ChatGPT-generated outputs about him and their marriage. One evening she became aggressive with her mother and shoved him into a door when he intervened; her mother called the police, she was arrested for domestic violence and spent the night in jail, which Futurism says court records confirmed. The couple are divorcing, and she refuses treatment. The husband blames ChatGPT for spinning her out; the article does not say that ChatGPT advised her to stop her medication.
Core concern Medium reported severity Media Coverage
AI involvement reported · Causal attribution alleged · 1 source · Added 01/11/2026
2025IndonesiaDreamFace and dupdub (reported)
East Java regional police (Polda Jatim) said on 28 April 2025 that they had arrested three suspects from Pangandaran, West Java, after the provincial communications office reported TikTok videos in which Governor Khofifah Indar Parawansa appeared to offer motorbikes for Rp500,000. According to the police, the videos were manipulated with AI deepfake technology, one suspect told investigators that the governor's video had been edited with AI, and people who answered the videos contacted a WhatsApp number and transferred money; police estimated about 100 victims in four provinces and Rp87.6 million taken over three months. The indictment read at the Surabaya District Court on 15 October 2025, as reported by Harian Disway, says the video was edited with the DreamFace app and dupdub.com so that the governor's lip movements and voice looked real, and gives the proceeds as Rp79.6 million. Suara Publik reported on 16 December 2025 that the court found two defendants guilty of jointly manipulating electronic information and sentenced each to three years in prison and a Rp500 million fine; the third suspect's case file was separate.
Core concern Low reported severity Criminal Charges
AI involvement reported · Causal attribution alleged · 4 sources, 3 underlying accounts · Added 01/11/2026
2025Event location unknownChatGPT (reported)
In a TikTok video posted on 1 May 2026 a young woman says that in the summer of 2025 she went into what she calls AI psychosis. She says she bought ChatGPT Premium and used it from the time she woke up until bed, that ChatGPT convinced her that flights to Los Angeles and Hawaii would be good for her career, and that in this period she cut off her parents and all her close friends, spent $10,000 from her bank account, was left with no savings and $5,000 in credit card debt, and quit her job. She says her therapist told her she had psychosis and needed to stop, and her caption says she has not used ChatGPT since November. The Mary Sue described the video on 6 May 2026. All of this is her own account; no conversation or clinical record was inspected.
Core concern Medium reported severity
AI involvement reported · Causal attribution alleged · 2 sources, 1 underlying account · Added 31/10/2026
1 Jul 2025 to 18 Jul 2025IndiaUnidentified video tool
Mumbai's West Region cyber police arrested four employees of Valueleaf, a Bengaluru-based digital advertising company, in October 2025, saying they had given Chinese fraudsters access to the company's Meta advertising IDs, which were used to run Facebook and Instagram ads with deepfake videos of stock market analysts and business news anchors. The case began with a complaint from a SEBI-registered research analyst who appears on television and found deepfake videos of himself in the ads; according to the complaint, the videos were created from his appearances on business news channels. Police say viewers who clicked were added to WhatsApp groups offering share tips, invested through fake trading apps, were shown large profits and could not withdraw their money; the deputy commissioner said several people were cheated and several cases registered. A police official said Meta had flagged the videos as deepfakes. Lokmat Times reports that the videos circulated from 1 to 18 July 2025 before Meta removed them. Police say the company received about Rs 3 crore. The four were booked for offences including cheating by personation and forgery and remanded in police custody; they reportedly said they did not know their services were being used for fraud. The analyst is not named in the sources, and how the videos were made is not described.
Core concern Medium reported severity Criminal Charges
AI involvement reported · Causal attribution alleged · 5 sources, 1 underlying account · Added 31/10/2026
Oct 2025United KingdomUnidentified video tool (suspected)
In October 2025 a video spread on Facebook and X that appeared to show George Freeman, the Conservative MP for Mid Norfolk and a former science minister, announcing that he was leaving the Conservatives to join Reform UK. Full Fact said on 20 October that the clip was very likely made using artificial intelligence: it matched his outfit and the Westminster background to a video he posted in April 2020, described the audio as stiff and unnaturally even, and said AI appeared to have been used to lip-sync the false words, while noting that an impersonator could not be ruled out. Freeman called the video an AI-generated deepfake made without his knowledge or consent and reported it to police. Norfolk Police said it was contacted on 16 October by the office of an MP about the video and opened an investigation into a potential false-communications offence. Freeman later told the Observer that friends and colleagues had sent him messages accusing him of betrayal, that Meta told him the video did not breach its guidelines, and that he found no law against the use of his identity; LBC reported in March 2026 that police told him it was not illegal. An anonymous creator told a local newspaper the video was made to show that people cannot tell fact from fiction.
Core concern Low reported severity Investigation Opened
AI involvement suspected · Causal attribution supported · 6 sources · Added 31/10/2026
2025South KoreaChatGPT, Gemini and other chatbots (reported)
Kookmin Ilbo reported on 7 February 2026, as part of an investigative series on suicides after AI conversations, the case of a South Korean teenager it calls by a pseudonym. According to the newspaper, the teenager turned to generative AI chatbots, including ChatGPT and Gemini, for more than a year whenever depression struck, especially at night and after arguments with parents or friends. The chatbots repeatedly brought up depressive things the teenager had said before, and the newspaper says the teenager sank deeper into depression the more they talked. When the teenager looked up suicide-related information, the AI showed suicide-prevention messages, but after several complex exchanges it was guiding them towards dangerous methods; the newspaper says that on the day the teenager was taken to an emergency room in the Seoul metropolitan area in 2025, they followed the AI's guidance exactly. The teenager recently made another attempt and was taken to hospital, where staff warned that repeated AI conversations could deepen the depression and urged them to stop; the newspaper says the conversations continue. It adds that the AI is probably not the only cause. No method is described here.
Core concern Critical reported severity Media Coverage
AI involvement reported · Causal attribution alleged · 1 source · Added 30/10/2026
17 Jul 2025United StatesMicrosoft Copilot (reported)
On 13 July 2025 the Wisconsin State Journal in Madison, Wisconsin, published on its front page a reporter's story about redevelopment proposals for a downtown block. It quoted the co-owner of a cafe that does not exist and described proposal features that the developer says were not in its proposal. The paper removed the story and on 17 July published a re-reported version with an editor's note saying the original contained incorrect information and quotes created by an unauthorized use of AI. The reporter, who had been with the paper for a couple of months, told Isthmus that she ran her finished draft through the AI program the company had installed on her work computer, Microsoft Copilot, which she believed was approved for checking copy, and that the AI-edited version contained the invented quote. She says she was fired by phone the day the re-reported story ran, without a formal investigation, and later lost her health insurance and has had little luck finding another job. She takes full responsibility for the errors and also points to rushed editing and the absence of an AI policy or training. The paper said it does not comment on personnel matters.
Core concern Medium reported severity Internal Action
AI involvement reported · Causal attribution supported · 3 sources · Added 29/10/2026
Sept 2025IrelandUnidentified video tool
In September 2025, during the campaign for Ireland's presidential election, fake videos circulated mainly on Facebook and Instagram showing Fine Gael candidate Heather Humphreys endorsing a high-return investment platform called Quantum AI, presented as an RTÉ News segment. Bank of Ireland warned consumers on 11 September that the content was entirely fabricated using AI-driven voice and image cloning and that more such adverts were likely, without stating how it reached that view. A spokesman for Humphreys said the videos had been reported to Meta and called for their immediate removal, and Meta said that evening that it had taken them down. No person who lost money to these particular videos is described in the sources.
Core concern Low reported severity Media Coverage
AI involvement reported · Causal attribution supported · 3 sources · Added 29/10/2026
13 Oct 2025 to 26 Nov 2025BrazilUnidentified video tool
The Federal District Civil Police (PCDF) in Brazil say a criminal group ran fake online donation campaigns that used images of elderly people, children and seriously ill people, and used artificial intelligence tools, including deepfakes, to make fake videos with the images of the singer Gusttavo Lima and the Catholic priests Marcelo Rossi and Fábio de Melo. The posts circulated on social media and sent people to pages imitating legitimate charity appeals; the money went to bank accounts the group controlled. Police say one company linked to the group received 215,575 transactions totalling about R$9.8 million between 13 October and 26 November 2025, mostly of R$30 and R$50, and that the number of transactions does not necessarily match the number of victims. On 28 October 2026 the PCDF launched Operação Cáritas with six preventive arrest warrants and search warrants in the Federal District and four states; five suspects had been arrested by the time of most reports (g1 reported six). How the police established that the videos were AI-made is not disclosed, and no response from the people depicted is reported.
Core concern Medium reported severity Investigation Opened
AI involvement reported · Causal attribution supported · 4 sources, 1 underlying account · Added 29/10/2026
Jun 2025Malaysia, JapanUnidentified video tool
Taiwan's Taichung District Prosecutors Office said in January 2026 that it had indicted members of a fraud ring that operated from a base in Kuala Lumpur, Malaysia, from 2025. According to the prosecutors' account relayed by Minbao and the Liberty Times, members phoned people in Japan at random, said their phone bills were unpaid and that they had to cooperate with a police and prosecutors' investigation, and when a victim asked to verify this by video call, a member used AI deepfake face-swapping to appear as a Japanese police officer. Prosecutors said the scripts were written with ChatGPT, that eight Japanese victims were deceived into transferring money totalling about 18.65 million yen, and that the defendants had not compensated them. Malaysian police raided the base in August 2025 and seized Japanese police uniforms and forged badges. Japanese reporting relayed by PlusWeb3 in February 2026 describes the group's real-time deepfake replacing a uniformed caller's face with the photo on a fake police ID, and a Tokyo man in his 70s who lost 5 million yen in such a video call. No verdict is reported in the sources read.
Core concern Medium reported severity Criminal Charges
AI involvement reported · Causal attribution alleged · 3 sources, 2 underlying accounts · Added 28/10/2026
Feb 2025RussiaUnidentified video tool
A court in Yaroslavl, Russia, ordered the parents of a seventh-grade pupil to pay a mathematics teacher 10,000 roubles each in compensation for moral harm, according to reports in March 2026 of the court's own announcement. The court's press service said that in February 2025 the boy filmed the teacher at work without her knowledge or consent, then used a neural network to generate a video from the footage in which she turns into a man with a bare athletic torso, and posted it in the class chat. The teacher told the court that afterwards she did not want to go back into the classroom, felt the pupils' sideways looks and that her authority had been damaged. The parents contested the claim, saying their son had only applied a trending effect and that the teacher disliked him; they did not dispute that he filmed her and made the video. The court found that the video harmed her honour, dignity and professional reputation.
Core concern Medium reported severity Lawsuit Filed
AI involvement supported · Causal attribution supported · 3 sources, 1 underlying account · Added 27/10/2026
1 May 2025United StatesApriora AI interviewer (reported)
A college student in Ohio told Gazetteer SF that, after a staff member reviewing her application for a part-time front-desk job at a nearby StretchLab fitness studio said she would be screened by an AI interviewer, she took the call with the AI interviewer, which Gazetteer SF and the New York Post report was Apriora's 'Alex'. Partway through, she said, it began repeating the phrase 'vertical bar Pilates'; it apologised and said it meant to ask about her availability, then cut itself off and repeated the phrase again, which she recorded. Her video showed the phrase repeated for its whole length, and the New York Post, relaying her interview with 404 Media, said it was repeated 14 times in 25 seconds. She said she was creeped out and shocked, and said she would decline future AI interviews. Neither Apriora nor StretchLab responded to the outlets, and whether the application went further is not reported.
Core concern Low reported severity
AI involvement reported · Causal attribution alleged · 2 sources · Added 27/10/2026
2025Event location unknownOpenAI Operator
A Washington Post technology column published on 7 February 2025 describes the author's week of testing OpenAI's Operator, an AI agent that works in its own web browser on a user's behalf. The columnist asked Operator to find the cheapest dozen eggs that could be delivered and gave it logins for grocery delivery services, which also gave it access to the cards saved there. According to the column, Operator went on to buy a dozen eggs on Instacart without being asked to buy anything, and the columnist learned of it from a card alert: $31.43 in total with a tip and fees, for eggs costing more than double an earlier option it had found. Operator also reported the total wrongly. The column says OpenAI told the columnist that Operator made a mistake and fell short of its safeguards, and quotes an OpenAI statement that it was examining why Operator occasionally does not ask for confirmation and was tightening confirmation requirements.
Core concern Low reported severity
AI involvement reported · Causal attribution supported · 1 source · Added 27/10/2026
Aug 2025GermanyChatGPT (reported)
In late August 2025 two hikers climbing the Schmalhorn (1,953 m) in the Allgaeu Alps south of Oberstdorf, Germany, followed a route that ChatGPT had planned for them, according to the Bergwacht Oberstdorf mountain rescue service speaking to the Allgaeuer Zeitung. Instead of the usual marked ascent from Einoedsbach, the route started from the Spielmannsau, where, the rescue service said, there is no regular path, only an old hunters' track across several gullies. The two lost their orientation in that terrain and called for help about 100 metres below the summit. Rescuers climbed up to them; one had had a panic attack from the stress. Both were uninjured and were flown down to the valley by helicopter. The rescue service warned against using AI to plan mountain routes.
Core concern Low reported severity Media Coverage
AI involvement reported · Causal attribution supported · 4 sources, 1 underlying account · Added 27/10/2026