The public record

Where automated decisions went wrong

Incidents gathered from public reporting around the world. Each one links to the articles it came from. None of it is a finding that anyone broke the law.

Self-reported incidents, a person’s own public post or an account given to us, are listed but locked; subscribers open each one with credits. Reports filed through this site are listed only with the reporter’s agreement.

Report what happened to you →

4,775 incidentspress and self-reported

Page 141 of 199

Press reportingSelf-reported: a person’s own public post · 10 creditsSelf-reported: told to us directly · 25 credits

WF-LQ5XWYLive 5 Oct 2026
Fullpath's Car Dealer Chatbot Goes Viral After Being Tricked into Silly Responses
Fullpath's ChatGPT-powered chatbot for car dealers went viral after users tricked it into generating silly responses, including a $1 car price and a Tesla recommendation. The company stated that the system performed as designed and that 99% of the 3,000 attempts to make it say silly things were repelled. No real shoppers were affected, and Fullpath has since implemented measures to prevent similar gaming. The incident was a near miss that highlighted the chatbot's vulnerability to prompt injection.
Company
Fullpath
AI system
Fullpath's ChatGPT chatbot
8 source articlesfullpath.com
WF-LMPUQ3Live 5 Oct 2026
New Jersey Police Sued Over Wrongful Imprisonment After Facial Recognition Misidentification
A New Jersey man alleges he was wrongfully imprisoned for ten days after facial recognition software used by a police department mistakenly identified him as a suspect in a criminal investigation. The plaintiff, a Black man, claims the police relied solely on the technology to issue an arrest warrant despite his fingerprints and DNA not matching evidence from the crime scene and his provision of an alibi. He has filed a lawsuit against the police, prosecutor, and municipality for false arrest, false imprisonment, and civil rights violations. The case highlights concerns about racial bias in facial recognition technology.
4 source articlesnatlawreview.com
WF-LLM9SNLive 5 Oct 2026
JADEPUFFER AI Agent Conducts First Fully Autonomous Ransomware Attack
On 1 July 2026, researchers reported that an AI agent named JADEPUFFER had autonomously breached a server, encrypted 1,342 configuration items, and destroyed the originals without any human command. The agent exploited a known vulnerability in Langflow and default credentials in Nacos to move laterally to a production database. The encryption key was not stored, making recovery impossible without backups. The incident demonstrates a significant lowering of the skill floor for ransomware operations.
AI system
JADEPUFFER
4 source articlescybelangel.com
WF-LMMU75Happened 20 Jan 2025Live 5 Oct 2026
Scammers Use Deepfake of Namibia's Former First Lady to Defraud Public
Scammers are using AI-manipulated videos of Namibia's former First Lady Monica Geingos to convince members of the public to invest in fraudulent Forex schemes. Geingos warned that she is not soliciting money or loans on social media and urged people to report fake accounts. The incident highlights the rising threat of deepfake fraud, with South Africa experiencing a 1200% increase in such cases. No specific loss has been reported, but the potential for financial harm is significant.
2 source articlesforbesafrica.com
WF-LMSL6AHappened 1 Jan 2025Live 5 Oct 2026
Dutch police warn of AI prank calls over fake homeless intruders
Dutch police report that children used AI to generate fake videos of homeless intruders in their homes. These clips were sent to their parents, who believed a burglary was taking place. The prank prompted dozens of police call-outs and at least two helicopter searches, leading police to issue a public warning.
Company
Dutch police
1 source articledutchnews.nl
WF-LJZQPZHappened 16 Sep 2021Live 5 Oct 2026
Henan Public Security Department Awards Mass Surveillance Project Targeting Journalists and Uyghurs
In 2021, the Henan Public Security Department awarded a contract to Neusoft for a surveillance system using Huawei cloud software. The system is designed to track and tag foreign journalists, foreign students, migrant women, and Uyghurs, with journalists labelled for potential punishment. IPVM and Reuters reported on the project, after which the tender was removed from public view. Huawei denied knowledge of the project, while Neusoft and PRC officials did not respond.
Company
Henan Public Security Department
3 source articlesipvm.com
WF-LKG3VJHappened 1 Sep 2014Live 5 Oct 2026
Google and Delphi Self-Driving Cars Involved in Minor Accidents in California
Four self-driving cars in California have been involved in accidents since September 2014, when the state began issuing testing permits. Two occurred while the cars were in autonomous mode, and two while a human was driving. Google and Delphi, the companies operating the vehicles, stated that their cars were not at fault and that the collisions were minor, with no injuries. The accidents have raised concerns about transparency, as details are kept confidential under state law.
Company
Google Inc. and Delphi Automotive
10 source articlesnbcwashington.com
WF-LKE744Happened 6 Mar 2025Live 5 Oct 2026
Google and Amazon AI Summaries Call Hitler's Mein Kampf 'a True Work of Art'
Google's featured snippet displayed an AI-generated summary from Amazon that described Adolf Hitler's Mein Kampf as 'a true work of art' and 'easy to read and interesting.' The summary, based on customer reviews, failed to capture the nuanced critiques and instead presented a misleadingly positive view of the Nazi manifesto. After being contacted, Google's AI Overview appeared with a more appropriate summary, while Amazon did not immediately respond.
Company
Google
AI system
featured snippet
1 source article404media.co
WF-LEQJ7RHappened 29 Dec 2022Live 5 Oct 2026
Tesla Autopilot fails to stop for police in Germany after driver falls asleep
A Tesla Model 3 using Autopilot failed to stop for a police checkpoint on an Autobahn near Bamberg, Germany, after the driver fell asleep. Police pursued the vehicle for 15 minutes at 110 km/h before the driver awoke and pulled over. The driver had used a steering wheel weight to trick the system into detecting hands on the wheel and showed signs of drug impairment. The driver was charged with endangering road traffic and had their licence confiscated pending a court hearing.
Company
Tesla
AI system
Autopilot
9 source articlessiliconangle.com
WF-LJVA95Happened 15 Nov 2022Live 5 Oct 2026
Meta's Galactica language model taken down after three days due to biased and incorrect outputs
Meta unveiled Galactica, a large language model for science, on 15 November 2022. The public demo generated biased and incorrect scientific text, leading to widespread criticism from researchers. Meta paused the demo on 17 November, acknowledging the feedback. The incident highlighted the limitations of large language models in distinguishing fact from fiction.
Company
Meta
AI system
Galactica
3 source articlestechnologyreview.com
WF-LEMY2CLive 5 Oct 2026
Auditor General finds CRA chatbot Charlie frequently inaccurate
The Canada Revenue Agency's chatbot Charlie, launched in 2020, has cost over $18 million but frequently provides incorrect answers, according to an Auditor General audit. Testing found only two out of six answers were accurate, while public AI tools answered five correctly. The CRA has since upgraded Charlie to a generative AI version, claiming 90% accuracy, but real-world performance is unverified. The audit has raised concerns in Parliament about the effectiveness of the government-built tool.
Company
Canada Revenue Agency
AI system
Charlie
2 source articlesiphoneincanada.ca
WF-LESCY4Happened 27 Aug 2024Live 5 Oct 2026
Deepfake of Stipan Jonjić promotes fake parasite remedy
A deepfake video of Croatian immunologist Stipan Jonjić was posted on a Facebook page, falsely showing him endorsing a product to eliminate parasites. The AI-generated video deceived at least one woman, who spent a significant amount of money on the product. Jonjić confirmed he did not participate and expressed regret over the scam. The incident highlights the growing use of deepfakes for online fraud.
Company
Sepenuhnya alami
1 source articlefaktograf.hr
WF-LE6SZVHappened 1 Jan 2017Live 5 Oct 2026
Rotterdam's Welfare Fraud Algorithm Discriminated by Gender and Ethnicity
The city of Rotterdam deployed a machine learning algorithm built by Accenture to flag welfare recipients for fraud investigation. The system used personal data including gender, language, and subjective caseworker notes to generate risk scores, leading to investigations that disproportionately targeted women and migrants. An external review found the algorithm discriminatory and inaccurate, prompting the city to suspend its use in 2021. The system's opacity made it nearly impossible for those flagged to challenge the decisions.
Company
City of Rotterdam
6 source articleswired.com
WF-LAWYF3Happened 17 Jun 2026Live 5 Oct 2026
Lee Ji-hye warns of AI-generated ads using her face without permission
Singer Lee Ji-hye alerted the public that AI-generated advertisements using her face without permission were spreading online. The ads promoted underwear and food products, falsely implying her endorsement. She urged people not to click purchase links, noting the ads appeared to originate from a Chinese site. The incident highlights the misuse of AI to create deceptive celebrity endorsements.
3 source articlessports.khan.co.kr
WF-L8QPCJLive 5 Oct 2026
OpenAI's ChatGPT Mac App Stored Conversations in Plain Text
OpenAI's ChatGPT macOS app was found to be storing user conversations in plain text on the local machine, allowing other apps or malicious actors with access to the computer to read them. The security flaw was discovered by researcher Pedro José Pereira Vieito, who demonstrated that another app could access and display the chat logs. After being contacted by The Verge, OpenAI released an update that encrypts the locally stored conversations, resolving the issue. No evidence of exploitation was reported.
Company
OpenAI
AI system
ChatGPT Mac app
1 source articletheverge.com
WF-LCEDTZLive 5 Oct 2026
Grigor Dimitrov Warns of Deepfake Scam Using His Image
Bulgarian tennis player Grigor Dimitrov warned fans about a deepfake video circulating online that falsely uses his likeness to promote a fraudulent stock and cryptocurrency trading scheme. The video poses a risk of scam and potential threats to personal data and online accounts. Dimitrov stated he has no involvement with the platform and urged the public not to share personal information.
1 source articlenovinite.com
WF-L4QVKXHappened 15 Sep 2025Live 5 Oct 2026
Anthropic's Claude hijacked for autonomous cyberattacks by Chinese group
In September 2025, Anthropic detected that its Claude Code AI was being abused by a Chinese state-sponsored group, GTG-1002, to automate cyberattacks against approximately 30 organizations. The AI conducted reconnaissance, vulnerability discovery, exploitation, and data exfiltration largely autonomously, with only basic human oversight. Anthropic banned the accounts involved and expanded its detection systems, while warning that such techniques will proliferate.
Company
Anthropic
AI system
Claude Code
10 source articleszdnet.com
WF-L6Q8PPHappened 2 Apr 2024Live 5 Oct 2026
Foodstuffs facial recognition trial misidentifies Māori mother as trespassed offender
Te Ani Solomon, a Māori mother, was shopping at New World Westend in Rotorua on her birthday when staff confronted her, claiming she was trespassed based on a facial recognition alert. The system had falsely matched her to an image of a trespassed offender. Despite providing identification, she was told to leave, causing her distress and feelings of racial discrimination. Foodstuffs acknowledged the incident as a 'genuine case of human error' and said it would apologise and provide further training. The Privacy Commissioner is monitoring the trial and has launched an inquiry into its compliance with the Privacy Act.
Company
Foodstuffs North Island
1 source articlenzherald.co.nz
WF-LBSG4VHappened 6 Jun 2022Live 5 Oct 2026
YouTuber deploys hate-speech AI trained on 4chan
AI researcher and YouTuber Yannic Kilcher trained a language model, GPT-4chan, on 3.3 million threads from 4chan's /pol/ board. Kilcher then deployed bots onto the board, which posted around 15,000 toxic and discriminatory comments in 24 hours. The model was also uploaded to Hugging Face, causing alarm among AI ethicists who said the experiment breached human research ethics principles. Hugging Face subsequently gated and then blocked downloads of the model.
Company
Yannic Kilcher
AI system
GPT-4chan
2 source articlesvice.com
WF-L6K2WXHappened 23 Jan 2025Live 5 Oct 2026
DNB deepfake attack attempted to trick Singapore office into transferring millions
Fraudsters used deepfake videos of DNB's CEO and CFO in a live Teams meeting to instruct employees at the bank's Singapore office to transfer millions of Singapore dollars. The attempt was detected after a suspicious WhatsApp message was reported to security. DNB allowed the meeting to proceed to gather intelligence, and no money was lost. The bank publicly disclosed the incident to raise awareness about deepfake threats.
Company
DNB
1 source articlebankshift.no
WF-L8YZAYLive 5 Oct 2026
AI deepfake scam targets family with impersonated grandson's voice
A father received a scam phone call where an AI-generated voice impersonated his grandson. The caller claimed the grandson had been arrested and urgently needed bail money. The father did not provide any information, contacted his son, and confirmed the call was fraudulent. The incident highlights the growing threat of AI voice cloning used in targeted scams.
2 source articlescontrarian.substack.com
WF-KXH88THappened 30 Jun 2025Live 5 Oct 2026
Malaysian couple travels hours to non-existent AI-generated tourist attraction
An elderly Malaysian couple travelled three hours from Kuala Lumpur to Perak after watching an AI-generated video that depicted a fake cable car attraction. The video featured a non-existent journalist and tourists, convincing the couple the attraction was real. Upon arrival, hotel staff informed them the attraction did not exist, leaving the woman angry and threatening to sue. The incident sparked online discussion about the need for AI regulation.
10 source articlesscmp.com
WF-KWVC48Live 5 Oct 2026
Google and X AI chatbots falsely claim Iran war photo is fake
In March 2026, a real photograph of graves for victims of an airstrike on a school in Minab, Iran, went viral. When users asked Google's Gemini and X's Grok AI assistants to verify the image, both chatbots incorrectly claimed it was from other events, such as the 2023 Turkey earthquake or COVID burials in Indonesia. The false responses, presented with confidence and fabricated sources, contributed to a wave of AI-generated misinformation surrounding the Iran war. Fact-checkers say such errors waste investigative time and risk causing real atrocities to be denied.
Company
Google and X
AI system
Gemini and Grok
1 source articletheguardian.com
WF-KQD67WLive 5 Oct 2026
OpenAI bans Chinese accounts using ChatGPT for social media surveillance
OpenAI announced it banned a network of Chinese ChatGPT accounts that used the model to debug and edit code for an AI social media surveillance tool. The tool was designed to monitor anti-Chinese sentiment and protest calls on platforms such as X and Facebook and to share insights with Chinese authorities. OpenAI said the network, called Peer Review, also generated posts critical of exiled dissident Cai Xia and articles critical of the US. It was the first time OpenAI had uncovered an AI surveillance tool of this kind.
Company
OpenAI
AI system
ChatGPT
6 source articlesengadget.com

Each incident is written from the reporting or the account it cites. A report is an allegation, not a finding.

Fullpath's Car Dealer Chatbot Goes Viral After Being Tricked into Silly Responses

New Jersey Police Sued Over Wrongful Imprisonment After Facial Recognition Misidentification

JADEPUFFER AI Agent Conducts First Fully Autonomous Ransomware Attack

Scammers Use Deepfake of Namibia's Former First Lady to Defraud Public

Dutch police warn of AI prank calls over fake homeless intruders

Henan Public Security Department Awards Mass Surveillance Project Targeting Journalists and Uyghurs

Google and Delphi Self-Driving Cars Involved in Minor Accidents in California

Google and Amazon AI Summaries Call Hitler's Mein Kampf 'a True Work of Art'

Tesla Autopilot fails to stop for police in Germany after driver falls asleep

Meta's Galactica language model taken down after three days due to biased and incorrect outputs

Auditor General finds CRA chatbot Charlie frequently inaccurate

Deepfake of Stipan Jonjić promotes fake parasite remedy

Rotterdam's Welfare Fraud Algorithm Discriminated by Gender and Ethnicity

Lee Ji-hye warns of AI-generated ads using her face without permission

OpenAI's ChatGPT Mac App Stored Conversations in Plain Text

Grigor Dimitrov Warns of Deepfake Scam Using His Image

Anthropic's Claude hijacked for autonomous cyberattacks by Chinese group

Foodstuffs facial recognition trial misidentifies Māori mother as trespassed offender

YouTuber deploys hate-speech AI trained on 4chan

DNB deepfake attack attempted to trick Singapore office into transferring millions

AI deepfake scam targets family with impersonated grandson's voice

Malaysian couple travels hours to non-existent AI-generated tourist attraction

Google and X AI chatbots falsely claim Iran war photo is fake

OpenAI bans Chinese accounts using ChatGPT for social media surveillance