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 →

4,442 incidentspress and self-reported

Page 74 of 186

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

WF-3GK2BUHappened 6 Dec 2023Live 5 Oct 2026
Lawyer used ChatGPT to cite fake cases in BC parenting application
In a British Columbia Supreme Court case, lawyer Chong Ke used ChatGPT to generate legal case citations for a notice of application in a parenting dispute. The citations were fictitious and were discovered by opposing counsel, who incurred costs investigating them. Ke later acknowledged the error, apologized, and withdrew the fake cases before the hearing. The court considered whether to order special costs against Ke personally.
AI system
ChatGPT
9 source articlesbccourts.ca
WF-HJZ02QHappened 1 Mar 2024Live 5 Oct 2026
Ukrainian YouTuber Olga Loek finds AI clones selling Russian goods on Chinese social media
Ukrainian YouTuber Olga Loiek discovered AI-generated clones of herself appearing on Chinese social media platforms such as HeyGen and Douyin, impersonating a Russian woman named Natasha, Anna, or Grace. The clones were used to sell Russian goods to China and to praise Russia-China relations without her consent. Loiek and her followers complained to the platforms, resulting in some videos being removed, but the issue remains largely unaddressed.
Company
Unknown (Chinese social media platforms HeyGen and Douyin)
AI system
Unnamed AI clone generation system
6 source articlesreadwrite.com
WF-7EKKSULive 5 Oct 2026
TurboTax and H&R Block AI chatbots give wrong tax advice
A Washington Post review found that AI chatbots from TurboTax and H&R Block were unhelpful or wrong up to half the time when answering tax questions. The chatbots are used by millions of taxpayers. The review tested the chatbots and found them to be unreliable.
Company
TurboTax and H&R Block
5 source articleswashingtonpost.com
WF-L05QOOLive 5 Oct 2026
Trump supporters create AI-generated fake images of black voters
Donald Trump supporters created and shared AI-generated fake images portraying black people as supporting the former president. The images were designed to encourage African Americans to vote Republican. One creator, radio host Mark Kaye, admitted he was not claiming accuracy. Another widely viewed image was posted by a satirical account. The BBC investigation found dozens of such deepfakes ahead of the US presidential election.
7 source articlesbbc.co.uk
WF-HT0OQCHappened 3 Oct 2023Live 5 Oct 2026
AI-generated bots disrupt Beaverton City Council meeting with hate speech
During a Beaverton City Council virtual meeting on October 3, 2023, over a dozen AI-generated speaker bots infiltrated the public comment period, spewing racist, anti-Semitic, and homophobic slurs. Mayor Lacey Beaty cut off the callers as the audience, including children, watched in shock. The council later acknowledged the incident and expressed empathy, stating it would work to improve processes to prevent recurrence. No group has claimed responsibility.
Company
Beaverton City Council
6 source articlesdailymail.co.uk
WF-4PQVAAHappened 1 Sep 2023Live 5 Oct 2026
Microsoft data centers in Arizona to use 50 million gallons of water annually for AI
Microsoft has built data centers in Goodyear, Arizona, to support AI workloads, including for OpenAI. The data centers use drinking water for evaporative cooling, with projected annual consumption of over 50 million gallons. This raises concerns about water scarcity in the desert region. Microsoft declined to comment on the assertion that the complex is partly designated for OpenAI's use.
Company
Microsoft
8 source articlestheatlantic.com
WF-PS9FKFHappened 1 Dec 2020Live 5 Oct 2026
Houston's ShotSpotter system delays police response and over-polices minority communities
Houston Police Department deployed ShotSpotter, a gunshot detection system, in Southeast and Northwest Houston starting in late 2020. The system alerts police to suspected gunfire, but over 80% of alerts are unfounded. This has led to longer response times for other calls and increased police presence in predominantly Black and brown neighborhoods, causing anxiety and concerns of over-policing. Critics argue the technology has not reduced gun violence and diverts resources from more effective strategies.
Company
Houston Police Department
AI system
ShotSpotter
7 source articleshoustonchronicle.com
WF-SJVPGFHappened 27 Feb 2024Live 5 Oct 2026
Microsoft Copilot generates fake press statements from Biden and Putin on Navalny's death
Microsoft's Copilot chatbot generated fake press statements attributed to US President Joe Biden and Russian President Vladimir Putin regarding the death of Russian opposition leader Alexei Navalny. A journalist using the tool for a news article received fabricated statements blaming Putin and a rebuttal. Microsoft investigated and stated it is making changes to improve response quality. Neither President actually issued the statements.
Company
Microsoft
AI system
Microsoft Copilot
10 source articleswindowscentral.com
WF-156KJJHappened 27 Feb 2024Live 5 Oct 2026
Study finds AI chatbots provide inaccurate election information
A study by AI Democracy Projects and Proof News found that AI chatbots from OpenAI, Meta, Google, Anthropic, and Mistral provided inaccurate election information more than half the time. The inaccuracies included false claims about voting methods and registration deadlines. The companies responded with varying explanations, and some plan to update their systems.
Company
OpenAI, Meta, Google, Anthropic, Mistral
AI system
ChatGPT-4, Llama 2, Gemini, Claude, Mixtral
7 source articlescbsnews.com
WF-Y1WBTUHappened 22 Mar 2022Live 5 Oct 2026
Citizens Advice finds ethnicity penalty in car insurance pricing
Citizens Advice conducted exploratory research into car insurance pricing and found that people of colour may be paying £250 more per year than White people. The research suggests that areas with large communities of colour may be identified as more risky by algorithms, even when objective risk factors are controlled. Citizens Advice has called on the Financial Conduct Authority to investigate the issue.
9 source articlescitizensadvice.org.uk
WF-T3IVP5Live 5 Oct 2026
Google News boosts AI-generated articles that rip off other outlets
Google News is boosting sites that use AI to rapidly churn out content that rips off other outlets, according to a 404 Media investigation. The article describes multiple examples of AI-generated articles appearing in Google News, including a site called Worldtimetodays.com that published a near-identical copy of a Distractify article with errors like 'war of stars fans' instead of 'Star Wars fans.' Google told 404 Media that it does not focus on whether an article was written by an AI or a human, opening the way for more AI-generated content on its platform.
Company
Google
AI system
Google News
10 source articles404media.co
WF-UDDP7BHappened 6 Dec 2017Live 5 Oct 2026
Illinois DCFS ends unreliable predictive analytics program for child abuse risk
The Illinois Department of Children and Family Services ended its Rapid Safety Feedback program, which used data mining to predict child abuse risk, after the agency's director called the technology unreliable. The system, developed by Eckerd Connects and Mindshare Technology, assigned risk scores to children but failed to flag several high-profile child deaths. The program was also criticised for overwhelming caseworkers with alerts and for potential bias against poor children of colour. DCFS decided not to renew the contract.
Company
Illinois Department of Children and Family Services
AI system
Rapid Safety Feedback
10 source articleschicagotribune.com
WF-LK6RRNHappened 24 Feb 2024Live 5 Oct 2026
Google Gemini chatbot gives biased response about PM Narendra Modi
Google's Gemini chatbot generated a biased and inappropriate response to a query about Indian Prime Minister Narendra Modi. A user on X shared screenshots, and Union minister Rajeev Chandrasekhar flagged the response as violating India's IT rules and criminal code. Google acknowledged the issue and said it had taken steps to address it, but warned that Gemini may not be reliable for current events and political topics.
Company
Google
AI system
Gemini
5 source articlestimesofindia.indiatimes.com
WF-9F92BDHappened 14 Nov 2019Live 5 Oct 2026
UK prison risk algorithm may embed racial bias, experts warn
The Ministry of Justice developed a digital tool to categorize prisoners, which experts and advocacy groups warn could automate and embed racism. The tool uses intelligence data that may reflect existing bias, potentially leading to disproportionate categorization of BAME prisoners. The MoJ claims no evidence of discrimination, but critics question the small trial sample and lack of transparency.
Company
Ministry of Justice
AI system
Digital prisoner categorisation tool
5 source articlesthebureauinvestigates.com
WF-51NEWFHappened 17 Feb 2024Live 5 Oct 2026
UIUC researchers weaponize GPT-4 to autonomously hack websites
Researchers at the University of Illinois Urbana-Champaign demonstrated that LLM-powered agents, particularly OpenAI's GPT-4, can autonomously hack vulnerable websites. In sandboxed tests, GPT-4 achieved a 73.3% success rate across five attempts on 15 vulnerabilities, while open-source models failed. The researchers used the OpenAI Assistants API, LangChain, and Playwright to enable the agents to interact with websites. The study highlights the potential for AI agents to be used in cyberattacks, with cost estimates suggesting they could be cheaper than human penetration testers.
Company
University of Illinois Urbana-Champaign
AI system
GPT-4
4 source articlestheregister.com
WF-WHUM5HLive 5 Oct 2026
Serco Leisure issued enforcement notices for unlawful biometric monitoring of employees
Serco Leisure and associated trusts used facial recognition and fingerprint scanning to monitor employee attendance at 38 leisure facilities. The ICO found they unlawfully processed biometric data of over 2,000 employees. Enforcement notices were issued ordering them to stop.
Company
Serco Leisure Operating Limited
8 source articlesico.org.uk
WF-ROMQCHHappened 5 Feb 2024Live 5 Oct 2026
OnlyFake site uses neural networks to generate fake IDs, bypasses OKX verification
An underground website called OnlyFake uses neural networks to generate realistic photos of fake IDs for $15. The journalist tested the service and obtained a convincing California driver's license. They then used another fake ID to successfully bypass the identity verification process on OKX, a cryptocurrency exchange. The article alleges that this technology could streamline bank fraud and money laundering, but reports no actual financial loss.
Company
OKX
AI system
OnlyFake
10 source articles404media.co
WF-T9JVOSHappened 26 Oct 2017Live 5 Oct 2026
Lyft approved driver with terrorism conviction due to background check failure
Lyft allowed a driver with a conviction for aiding terrorism to drive for the ride-sharing service. The driver, Raja L. Khan, had served seven years and was denied by Uber and the city of Chicago, but Lyft's independent background check provider approved him. After WGN Investigates reported the issue, Lyft deactivated the driver and apologised, stating the approval was unacceptable and that it would re-run background checks.
Company
Lyft
7 source articlespatch.com
WF-82DEW0Happened 1 Jan 2024Live 5 Oct 2026
Gab launches AI chatbots denying Holocaust and promoting Nazi ideology
The far-right social media company Gab launched AI chatbots, including one named 'Uncle A' that impersonates Adolf Hitler and denies the Holocaust. The chatbots, built on open source models, are fine-tuned to promote antisemitic and white supremacist views without censorship. Gab's CEO Andrew Torba defended the tools as a free speech alternative to mainstream AI.
Company
Gab
AI system
Based AI
8 source articlesrollingstone.com
WF-YQBC6GHappened 22 Feb 2024Live 5 Oct 2026
Vending machine facial recognition error message concerns University of Waterloo students
A vending machine at the University of Waterloo displayed an error message referencing a facial recognition app, sparking privacy concerns among students. The incident raised questions about whether the machines were collecting biometric data without consent. No further action or response has been reported.
Company
University of Waterloo
8 source articleskitchener.ctvnews.ca
WF-7VB2DRHappened 21 Feb 2024Live 5 Oct 2026
User reports AI image generator producing inaccurate historical images
A user interacting with an AI image generation system found that it produced images of historical figures that were not representative of the actual participants. The system apologised for deviating from the prompt and offered to create images adhering to the traditional definition, but the user's subsequent prompts did not yield satisfactory results. The user shared the interaction on social media, highlighting the system's failure to accurately depict historical events.
10 source articlestwitter.com
WF-NFJ9W9Happened 12 Feb 2024Live 5 Oct 2026
Frontiers retracts study with AI-generated rat penis images
A study published in the journal Frontiers in Cell Development and Biology was retracted after it was discovered that the authors used the AI tool Midjourney to generate nonsensical images, including a giant rat penis. The journal apologised and said a reviewer's concerns were ignored. The incident highlights issues with AI use in academic publishing.
AI system
Midjourney
8 source articlesvice.com
WF-JIK7EXHappened 7 Nov 2023Live 5 Oct 2026
Robot crushes worker to death at paprika sorting facility in South Korea
A robot at a paprika sorting facility in Goseong, South Korea, mistook a worker for a box and crushed him. The worker, a 40-year-old employee of the installation company, was checking the robot's sensors when it picked him up and compressed him. He died from his injuries. The facility operator expressed concern about safety and the need for more precise systems.
Company
Donggoseong Export Agricultural Complex
9 source articlesyna.co.kr
WF-UVK6J6Happened 1 Jan 2015Live 5 Oct 2026
Mattel discontinues Hello Barbie doll, service no longer works
Mattel has discontinued the Hello Barbie doll, an AI-powered toy that used speech recognition and WiFi to chat with children. The service for the doll is no longer working, leaving the toy non-functional. Mattel provides customer support contact information for affected consumers.
Company
Mattel
AI system
Hello Barbie doll
10 source articleshellobarbiefaq.mattel.com

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

Lawyer used ChatGPT to cite fake cases in BC parenting application

Ukrainian YouTuber Olga Loek finds AI clones selling Russian goods on Chinese social media

TurboTax and H&R Block AI chatbots give wrong tax advice

Trump supporters create AI-generated fake images of black voters

AI-generated bots disrupt Beaverton City Council meeting with hate speech

Microsoft data centers in Arizona to use 50 million gallons of water annually for AI

Houston's ShotSpotter system delays police response and over-polices minority communities

Microsoft Copilot generates fake press statements from Biden and Putin on Navalny's death

Study finds AI chatbots provide inaccurate election information

Citizens Advice finds ethnicity penalty in car insurance pricing

Google News boosts AI-generated articles that rip off other outlets

Illinois DCFS ends unreliable predictive analytics program for child abuse risk

Google Gemini chatbot gives biased response about PM Narendra Modi

UK prison risk algorithm may embed racial bias, experts warn

UIUC researchers weaponize GPT-4 to autonomously hack websites

Serco Leisure issued enforcement notices for unlawful biometric monitoring of employees

OnlyFake site uses neural networks to generate fake IDs, bypasses OKX verification

Lyft approved driver with terrorism conviction due to background check failure

Gab launches AI chatbots denying Holocaust and promoting Nazi ideology

Vending machine facial recognition error message concerns University of Waterloo students

User reports AI image generator producing inaccurate historical images

Frontiers retracts study with AI-generated rat penis images

Robot crushes worker to death at paprika sorting facility in South Korea

Mattel discontinues Hello Barbie doll, service no longer works