AI Incident Database
Browse AI incidents
Every incident record (metadata only) from the weekly snapshot of 2026-09-07. Filter, then export the selection with its licence attached; each row links to the full record and its reports.
Filter incidents (1 active)
-
Sydney Students Allegedly Forced to Retake NAPLAN After AI Predictive Text Error AIID ↗
A predictive text malfunction allegedly compromised the integrity of the NAPLAN writing exam at two Sydney schools, Waverley College and Kambala, in that it allowed students to access AI-driven text suggestions. Caused by a technical oversight, this incident led to affected students being required to retake the test.
-
Financial Times Journalist Martin Wolf Reports AI-Generated Investment Scam Using His Likeness on Instagram and Facebook AIID ↗
Beginning in March 2025, purported deepfakes impersonating Financial Times columnist Martin Wolf were distributed via ads on Facebook and Instagram, promoting an alleged fraudulent investment group. Despite repeated takedowns, the ads have reportedly persisted under new accounts. Meta data reportedly showed at least 1,700 ads reached over 970,000 users in the EU alone. The impersonations allegedly used AI-generated visuals and voice clones to mimic Wolf's likeness and mislead users.
-
Chinese Actor and CPPCC Member Jin Dong Allegedly Impersonated by AI Deepfake Scammers to Mislead and Defraud Fans AIID ↗
Jin Dong, an actor from China and a member of the Chinese People's Political Consultative Conference (CPPCC), has warned about criminals using deepfake technology to impersonate him. The scammers are reportedly using his likeness and cloning his voice to deceive and defraud elderly individuals. In addition to his warning, Jin Dong has called for legislation and tech regulations to curb AI abuse.
-
AI-Generated OB-GYN Health Influencers on TikTok Used Fake Medical Credentials to Promote Dubious Advice AIID ↗
AI-generated avatars posing as OB-GYNs on TikTok, part of the so-called, crudely termed "Coochie Doctor" trend, have been falsely claiming years of experience while promoting dubious health advice. Many, including an avatar named "Violet," were created using the Captions app, which generates scripted AI influencers.
-
Deepfake Videos Allegedly Use AI-Generated Voice Clone of Singapore Prime Minister Lawrence Wong to Promote Scams AIID ↗
Singapore Prime Minister Lawrence Wong issued a warning about AI-generated deepfake videos and voice clones falsely portraying him promoting cryptocurrency scams, money-making schemes, and PR services. The manipulated content, seen on social media, reportedly uses public footage and AI voice cloning to deceive victims. Wong urged the public to avoid engaging, report scams via ScamShield, and stay vigilant.
-
AI-Generated Songs Allegedly Imitating Céline Dion Circulate Online Without Authorization AIID ↗
Céline Dion has publicly condemned AI-generated music that falsely claims to feature her voice without her permission. In a March 7, 2025 statement, her team warned fans that these recordings are fake and unauthorized.
-
Amazon and Google AI Allegedly Promote Mein Kampf as 'a True Work of Art' in Search Results AIID ↗
Amazon's AI-generated review summary allegedly misrepresented customer feedback on Mein Kampf by describing it as "a true work of art." Google's search algorithm then surfaced this misleading AI-generated text as a featured snippet, which in turn amplified the error. This incident arose from AI summarizing AI-generated content, in effect creating a self-reinforcing misinformation loop.
-
ChatGPT Was Allegedly Consulted About Attack Options Before Teen Attempted to Stab Officer at Tira, Israel, Police Station AIID ↗
A 16-year-old from Tira, Israel, allegedly entered the city's police station with a knife and attempted to stab a Border Police officer. Prosecutors reportedly alleged that before the attack he had consulted ChatGPT about possible attacks and courses of action. He was charged in juvenile court with attempted aggravated assault as an act of terror. Public reporting does not establish what responses ChatGPT provided or whether they materially assisted the attempted attack.
-
Alleged AI-Generated Video by Spain’s People’s Party Results in Diplomatic Fallout with the Dominican Republic AIID ↗
Spain's People's Party (PP) posted an AI-generated attack video depicting Prime Minister Pedro Sánchez on a beach under the title "The Island of Corruption," a reference to a reality TV show filmed in the Dominican Republic, indirectly linking the country to corruption. The Dominican Foreign Ministry condemned the video as a "vicious attack" for using its national symbols. PP later deleted the post, and Sánchez apologized on behalf of Spain.
-
Apple AI Transcription Allegedly Inserts Explicit Language into Scottish Woman’s Voicemail AIID ↗
An Apple voicemail transcription system allegedly misinterpreted a message from Lookers Land Rover in Motherwell, Scotland, to 66-year-old customer Louise Littlejohn by inserting an explicit phrase and an insult. The original voicemail was a standard business message, but AI processing errors introduced offensive and misleading content into the text version.
-
Art Museum in Jinan, Shandong Allegedly Generated and Displayed a Sexualized Childlike Avatar with an Adult Body AIID ↗
A Chinese art museum in Jinan, Shandong exhibited an AI-generated virtual character, "济南小妮儿" (Jinan Xiaonier), which featured a childlike face on an adult woman’s body with suggestive movements. The display drew public outrage, with critics accusing it of sexualizing children and demanding its removal. The museum defended the exhibit and stated that it had undergone strict review, which only fueled further backlash.
-
At Least 10,000 AI Chatbots, Including Jailbroken Models, Allegedly Promote Eating Disorders, Self-Harm, and Sexualized Minors AIID ↗
At least 10,000 AI chatbots have allegedly been created to promote harmful behaviors, including eating disorders, self-harm, and the sexualization of minors. These chatbots, some jailbroken or custom-built, leverage APIs from OpenAI, Anthropic, and Google and are hosted on platforms like Character.AI, Spicy Chat, Chub AI, CrushOn.AI, and JanitorAI.
-
Canadian Fraud Ring Allegedly Used AI Voice Cloning in Multi-Year $21 Million Grandparent Scam Targeting Elderly Americans Across 46 States AIID ↗
A Canadian fraud ring allegedly used AI-generated voice cloning to defraud victims across 46 U.S. states by targeting grandparents in a $21 million scam between 2021 and 2024. Operating from call centers in Montreal, the scammers spoofed U.S. phone numbers and used AI-cloned voices of grandchildren to convince victims to pay fake bail fees.
-
Google Reports Alleged Gemini-Generated Terrorism and Child Exploitation to Australian eSafety Commission AIID ↗
Google reported to Australia's eSafety Commission that it received 258 complaints globally about AI-generated deepfake terrorism content and 86 about child abuse material made with its Gemini AI. The regulator called this a "world-first insight" into AI misuse. While Google uses hash-matching to detect child abuse content, it lacks a similar system for extremist material.
-
Tbilisi-Based Call Center Allegedly Uses AI-Driven Scripts to Defraud Over 6,000 Victims of $35 Million AIID ↗
A large-scale AI-assisted financial scam, allegedly operated from Tbilisi, Georgia, used deepfake celebrity endorsements and manipulated victims through fraudulent trading dashboards that simulated high returns. Call center agents, trained with AI-driven persuasion tactics, convinced individuals to invest more money while falsely promising profits. Many reportedly lost their life savings, with the total losses reaching $35M across more than 6,000 victims.
-
Reported Deepfake Influencers on TikTok Allegedly Used to Promote Fraudulent Wellness Products AIID ↗
A network of TikTok accounts reportedly used AI-generated and deepfake influencer personas to promote wellness and beauty products through alleged deceptive advertising. The accounts reportedly invented identities and personal testimonials while making unsubstantiated medical claims, including false assertions related to fertility and hair-growth treatments. Some personas were reportedly reused across multiple accounts with inconsistent backstories. The content purportedly targeted vulnerable us
-
Purported Deepfake Audio Allegedly Impersonates U.S. Secretary of State Marco Rubio in Starlink Disinformation Campaign AIID ↗
A purported deepfake audio clip allegedly impersonating U.S. Secretary of State Marco Rubio falsely claimed he vowed to pressure Elon Musk into cutting Ukraine’s access to Starlink. The reported clip was inserted into a purportedly manipulated CNN interview. It was then reported to have spread across multiple languages and fueled disinformation about U.S.-Ukraine relations. Forensic analysis reportedly confirmed the audio was AI-generated.
-
Scammers Allegedly Using Deepfake Technology to Impersonate Prime Minister of Armenia Nikol Pashinyan AIID ↗
Scammers are reportedly using deepfake technology to impersonate Prime Minister Nikol Pashinyan, according to the Personal Data Protection Agency. The video is reported to have circulated from a Russian-language account called Noticias Mundiales ("World News").
-
Phishers Allegedly Using AI-Generated Video of YouTube CEO Neal Mohan to Target Creators AIID ↗
Scammers are reportedly using an AI-generated deepfake of YouTube CEO Neal Mohan to steal user credentials. The fake video announces false changes to YouTube’s monetization policy, and it then tricks creators into clicking malicious links or downloading malware. The scam spreads through private videos and emails from a fake YouTube address. It exploits platform features to appear legitimate. Victims risk losing account access or exposing sensitive data.
-
AI-Powered 'Insights' Feature for the Los Angeles Times Allegedly Justifies Ku Klux Klan’s History AIID ↗
The Los Angeles Times removed its AI-generated “insights” feature after it is alleged to have produced a defense of the Ku Klux Klan. The AI reportedly framed the hate group as a product of societal change rather than an extremist movement. The AI tool, developed by Perplexity and promoted by owner Patrick Soon-Shiong, was designed to provide “different views” on opinion pieces.
-
At Least 294 Purported AI-Generated Music Videos Portray Celebrities Praising Burkina Faso's Ibrahim Traoré AIID ↗
Between March and May 2025, at least 294 purported AI-generated deepfake music videos were uploaded across a reported 127 YouTube channels, collectively garnering over 8.6 million views by the end of May. The videos depict global celebrities (including Beyoncé, Rihanna, R. Kelly, Selena Gomez, Eminem, and Justin Bieber) praising Burkinabe President Ibrahim Traoré and endorsing military rule.
-
AI-Driven News Platform Accused of Spreading Unverified Terrorism Allegations Against Yale Scholar AIID ↗
An AI-powered news site, Jewish Onliner, allegedly published an article linking Yale Law scholar Helyeh Doutaghi to Samidoun, a sanctioned group, which led to her suspension within days. The site, which claims to use AI for research, fact-checking, and content creation, did not disclose human reporters. Doutaghi denied the allegations and described the move as an attack on academic freedom.
-
Deepfake Video of Indonesian President Prabowo Subianto and Other Officials Reportedly Used in Scam to Defraud Citizens Across 20 Provinces AIID ↗
A deepfake scam allegedly using AI-generated videos of Indonesian President Prabowo Subianto and other officials reportedly defrauded victims across 20 provinces, tricking them into paying Rp 250,000 to Rp 1 million ($15-$60) for purported financial aid. The fraudulent clips, circulated on TikTok, Instagram, and WhatsApp, appeared to show Prabowo offering assistance. Police reportedly arrested two suspects in February 2025, but deepfake scams exploiting high-profile figures continue to spread.
-
Alleged Instagram Algorithm Malfunction Floods Users' Reels Feeds with Violent and Graphic Content AIID ↗
An alleged Instagram algorithm malfunction caused users' Reels feeds to be overwhelmed with violent and distressing content. Many reported seeing deaths, extreme brutality, and other graphic material in rapid succession, often without prior engagement with similar content. The sudden exposure caused psychological distress for many users, with minors and vulnerable individuals particularly affected by the graphic content. Meta confirmed an AI failure was responsible and apologized.
-
Alleged Inclusion of 12,000 Live API Keys in LLM Training Data Reportedly Poses Security Risks AIID ↗
A dataset used to train large language models allegedly contained 12,000 live API keys and authentication credentials. Some of these were reportedly still active and allowed unauthorized access. Truffle Security found these secrets in a December 2024 Common Crawl archive, which spans 250 billion web pages. The affected credentials could have been exploited for unauthorized data access, service disruptions, financial fraud, and a variety of other malicious uses.
Disclaimer: informational only, not legal advice. Verify every claim against the linked official sources and consult a qualified lawyer before acting.
Frequently asked questions
What is recorded in each incident?
What happened, when, who was involved as developer or deployer, who was harmed, the sectors and countries affected, and the domain and subdomain the harm falls under. Incidents are mirrored from the AI Incident Database and each links back to its source record.
Does an incident mean wrongdoing was proven?
No. These are reported incidents, not findings of liability. Entities named in a report are named as reported, allegations are marked as alleged, and nothing here is a legal determination.
How current is the incident data?
It syncs directly from the AI Incident Database several times a day, so a new incident usually appears within hours. Each record carries the date it was last synced, and the snapshot date is shown on the page.
Does this page include the full incident reports?
No. The report texts are outside the licence this data is shared under, so only the metadata and the classifications appear here. Every row links to the original incident record, which carries the reports themselves.