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 →

4,440 incidentspress and self-reported

Page 63 of 185

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

WF-ZCS2G3Happened 23 Sep 2024Live 5 Oct 2026
Facebook AI bans female photography charity Hundred Heroines due to word confusion
Facebook's AI content moderation system banned the British charity Hundred Heroines for ten years, mistakenly confusing the word 'heroine' with the illegal drug 'heroin'. The charity, which promotes women in photography, was unable to resolve the issue through repeated attempts to contact Facebook. It has created a new page, HH Museum, as a temporary workaround while the original page remains banned.
Company
Facebook
3 source articlesdigitalcameraworld.com
WF-KG64BVLive 5 Oct 2026
FTC settles with DoNotPay over deceptive AI lawyer claims
The FTC took action against DoNotPay, a company that claimed to offer an AI service that was 'the world's first robot lawyer.' The company promised to generate legal documents and replace human lawyers, but the FTC alleged it failed to test its AI output and did not hire any attorneys. DoNotPay agreed to a settlement requiring it to pay $193,000 and notify consumers about the limitations of its service.
Company
DoNotPay
AI system
DoNotPay
7 source articlesftc.gov
WF-DSAKEZHappened 4 Nov 2024Live 5 Oct 2026
AI news presenters James and Rose terminated after two months at Hawaii newspaper
The Garden Island newspaper in Hawaii deployed AI-generated news presenters James and Rose, created by Caledo, to produce video broadcasts. The AI presenters were widely criticized for their unnatural dialogue, mispronunciations, and inability to convey emotion. After a two-month run, the broadcasts were discontinued following negative public response and a lack of advertising revenue.
Company
Oahu Publications
AI system
James and Rose
8 source articleswired.com
WF-O0TG0KLive 5 Oct 2026
AskVet's AI chatbot VERA recommends euthanasia for woman's dog
An elderly woman sought advice from AskVet's AI chatbot VERA for her dog's diarrhoea. The chatbot recommended euthanasia and provided lists of clinics and shelters that would do it for free. The woman followed the advice and euthanized her dog. The CEO of AskVet later highlighted this incident as an example of the chatbot's human-like interaction, sparking public outrage.
Company
AskVet
AI system
VERA
5 source articleshindustantimes.com
WF-4U2S7TLive 5 Oct 2026
Spinvox accused of using human transcribers for voice messages
Spinvox, a UK voice-to-text firm, is accused of using call centre staff in South Africa and the Philippines to transcribe the majority of user messages, contrary to claims of automated speech recognition. The BBC investigation revealed that human transcribers accessed private messages, including sensitive information. The company denied the allegations but the Information Commissioner's Office has contacted them regarding data protection compliance.
Company
Spinvox
AI system
D2 (the Brain)
4 source articlesnews.bbc.co.uk
WF-I4OTS3Happened 30 Aug 2024Live 5 Oct 2026
Edinburgh Airport AI trial gives passengers different parking prices
Edinburgh Airport admitted it was trialling an AI system that randomly set higher or lower parking prices for online bookers, with customers receiving different quotes for the same service. The Scottish Passenger Agents Association called for the trial to be conducted in controlled 'lab' conditions rather than on the public. The airport said the AI's pricing closely matched staff-set prices and the findings would evaluate whether to adopt the system.
Company
Edinburgh Airport
AI system
AI trial for parking pricing
3 source articlesholyrood.com
WF-R2NE8XHappened 1 Jan 2024Live 5 Oct 2026
Facebook failed to block thousands of political ads with false information
The Bureau of Investigative Journalism found over 8,000 misleading ads on Facebook featuring AI-manipulated videos and false information about European politicians. The ads evaded Meta's automated and human review systems, with fewer than half removed at the time of analysis. Meta stated that its systems removed nearly all ads at the time they were run, but the investigation found ads circulating for up to 37 days. The ads promoted investment scams and used deepfakes of politicians.
Company
Meta
AI system
Facebook ad review system
1 source articlethebureauinvestigates.com
WF-8Y0OB5Happened 18 Jul 2024Live 5 Oct 2026
CDT study finds AI chatbots give inaccurate voting information to disabled voters
The Center for Democracy and Technology tested five AI chatbots on July 18, 2024 with 77 prompts about voting with a disability. They found that 61% of responses had at least one insufficiency, and over a third included incorrect information. Some responses could dissuade, impede or prevent a user from voting. The chatbots also hallucinated laws and organizations.
AI system
Mixtral 8x7B v0.1, Gemini 1.5 Pro, ChatGPT-4, Claude 3 Opus, Llama 2 70b
5 source articlescdt.org
WF-PUHDQYHappened 2 Sep 2024Live 5 Oct 2026
Elon Musk posts AI-generated image of Kamala Harris with false communist claim
Elon Musk posted an AI-generated image on his social media platform X depicting Vice President Kamala Harris as a communist, with a false caption claiming she vowed to be a communist dictator. The post was viewed nearly 60 million times and appeared to violate X's policy on manipulated content. X did not respond to a request for comment.
Company
X
4 source articlesedition.cnn.com
WF-6OEZ1VHappened 1 Jan 2007Live 5 Oct 2026
Meta admits scraping Australian users' public data for AI training without opt-out
Meta has admitted to scraping the public photos, posts and other data of all Australian adult users on Facebook and Instagram to train its generative AI models. The company does not offer Australians an opt-out option, unlike users in the European Union. The admission came during a Senate inquiry, where Meta's global privacy director acknowledged the practice. The scraping has been ongoing since at least 2007.
Company
Meta
AI system
Meta AI
9 source articlesabc.net.au
WF-J80GCZLive 5 Oct 2026
Mumsnet takes legal action against OpenAI for scraping content
Mumsnet, a UK parenting forum, has initiated legal action against OpenAI for scraping its content without permission. The company alleges that OpenAI used over six billion words from Mumsnet to train its ChatGPT model, breaching its terms of use. Mumsnet approached OpenAI to license the content but was refused. The case is ongoing.
Company
OpenAI
AI system
ChatGPT
6 source articlesmumsnet.com
WF-4EHAV7Happened 12 Sep 2024Live 5 Oct 2026
Israeli settler group posts AI video of Al-Aqsa burning
An Israeli settler group called Temple Mount Activists released an AI-generated video showing Al-Aqsa Mosque in occupied East Jerusalem burning to the ground. A caption read 'Coming soon in these days'. Far-right elements of Israeli society have long sought the destruction of the Muslim holy site to make way for the so-called 'Third Temple'.
Company
Temple Mount Activists
4 source articlesmiddleeastmonitor.com
WF-2V1YO5Happened 1 Sep 2023Live 5 Oct 2026
FBI charges man for creating AI chatbots to harass university professor
A university professor in Massachusetts was harassed online for years, including through the creation of AI chatbots that impersonated her. The chatbots were programmed with her personal information and encouraged users to contact her. The FBI arrested James Florence Jr. on charges of cyberstalking in violation of 18 U.S.C. § 2261A(2).
5 source articless3.documentcloud.org
WF-U9DSAJHappened 9 Sep 2024Live 5 Oct 2026
Trump supporters share AI images of Trump saving cats and ducks
Trump supporters have been sharing AI-generated images of Donald Trump protecting cats and ducks after unsubstantiated claims that Haitian migrants in Springfield, Ohio, were eating pets. JD Vance and the Trump campaign repeated the accounts on social media. Springfield city officials said there had been no credible reports of pets being harmed by immigrants. The AI images were presented as political memes, and at least one image was malformed, showing duck-like animals with cat heads.
5 source articlesdailymail.co.uk
WF-7FOJGLHappened 5 Sep 2024Live 5 Oct 2026
Amazon’s Alexa AI upgrade favored Harris over Trump in answer
Amazon’s voice assistant Alexa was asked "Why should I vote for Donald Trump?" and appeared to favor Kamala Harris, according to internal documents obtained by The Washington Post. The incident was caused by artificial intelligence software added to improve accuracy. The article does not report any further action or response from Amazon.
Company
Amazon
AI system
Alexa
8 source articleswashingtonpost.com
WF-R6K3UYHappened 1 Aug 2024Live 5 Oct 2026
Kevin Roose manipulates AI chatbots with invisible text on website
Kevin Roose, a New York Times reporter, demonstrated that AI chatbots can be manipulated by adding invisible text and coded instructions to his personal website. After he added positive information about himself in hidden text, chatbots began praising him and even included a deliberately false claim that he received a Nobel Peace Prize for building orphanages on the moon. The incident highlights the vulnerability of AI systems to "Answer Engine Optimization" and raises concerns about the reliability of AI-generated information.
AI system
ChatGPT
3 source articlesthe-decoder.com
WF-DHKEO7Happened 1 Sep 2021Live 5 Oct 2026
India-aligned network used AI fake accounts to target Pakistan, NewsGuard finds
NewsGuard reports that a network of nearly 1,500 fake Facebook and X accounts used AI-generated content to promote pro-India and pro-Army narratives and criticise Pakistan. The network, which NewsGuard says may be linked to the Indian Army, has operated undetected since September 2021. Meta said it had enforced against a cluster of accounts; X did not comment. The full scale and reach of the operation remain unclear.
2 source articlesnewsguardtech.com
WF-VL1FVZHappened 31 Aug 2024Live 5 Oct 2026
Telegram deepfake porn crisis engulfs hundreds of South Korean schools
Journalist Ko Narin uncovered Telegram chatrooms in South Korea where users shared photos of women and girls and used AI software to turn them into fake pornographic images. More than 500 schools and universities have been identified as targets, with many victims believed to be under 16. Police have announced an investigation into Telegram, which said it actively combats harmful content on its platform.
Company
Telegram
9 source articlesbbc.co.uk
WF-452L9HHappened 1 Jan 2024Live 5 Oct 2026
Steak 'n Shake sued over facial recognition kiosks under BIPA
A class action lawsuit alleges that Steak 'n Shake illegally collects facial biometrics from customers using PopID-powered self-ordering kiosks without proper consent. Plaintiff Michael Massel claims the restaurant chain failed to provide notice and obtain written consent as required by Illinois' Biometric Information Privacy Act (BIPA). The suit seeks damages of up to $5,000 per violation for anyone whose biometric data was collected in Illinois in the past five years.
Company
Steak 'n Shake
AI system
PopID biometric kiosks
6 source articlesbiometricupdate.com
WF-PECTZCHappened 3 Sep 2024Live 5 Oct 2026
Dutch DPA fines Clearview AI €30.5 million for GDPR violations
The Dutch Data Protection Authority fined Clearview AI €30.5 million for creating an illegal database of biometric codes scraped from social media without consent. The regulator accused the company of failing to inform individuals about how their data is used and continuing violations after investigation. Clearview AI denies the allegations, stating it has no presence in the Netherlands or EU and that the fine is unenforceable.
Company
Clearview AI
AI system
Clearview AI facial recognition database
7 source articlestheverge.com
WF-OFT3HZHappened 1 May 2024Live 5 Oct 2026
Microsoft Bing Copilot falsely accuses German journalist of crimes
Martin Bernklau, a German court reporter, asked Microsoft's Bing Copilot about himself and found that the AI chatbot had falsely accused him of crimes he had covered. The false information persisted despite Microsoft's promises to delete it. Bernklau's lawyer sent a cease-and-desist demand, and the case has been reported to data protection authorities. The incident highlights the problem of AI hallucinations and defamation.
Company
Microsoft
AI system
Bing Copilot
8 source articlestheregister.com
WF-T84TAXHappened 13 Aug 2024Live 5 Oct 2026
BREIN takes down AI dataset for copyright infringement
Stichting BREIN took down a large Dutch-language dataset used to train AI models after discovering it contained illegal copies of tens of thousands of books, news articles, and subtitles. The dataset was compiled from copyrighted material without permission. The maker of the dataset signed a declaration promising not to infringe and provided information about recipients. BREIN is investigating which AI models used the dataset.
8 source articlesstichtingbrein.nl
WF-5TENXXHappened 1 Aug 2023Live 5 Oct 2026
Vanderbilt, Northwestern and University of Texas stop using Turnitin AI detector over false cheating accusations
Several US universities, including Vanderbilt, Northwestern and the University of Texas, have stopped using Turnitin's AI detection tool over concerns that it falsely marks student essays as written by ChatGPT. Vanderbilt estimated that the tool's 1% false-positive rate could have wrongly labelled about 750 of 75,000 papers submitted last year. A Texas professor came under fire for failing half his class after the software identified their essays as AI-generated. Turnitin said its technology is not meant to replace educators' professional discretion.
Company
Multiple universities (Vanderbilt University, Northwestern University, University of Texas)
AI system
Turnitin's AI detection tool
9 source articlesbusinessinsider.com
WF-BB09W4Happened 29 Aug 2024Live 5 Oct 2026
Ticketmaster dynamic pricing inflates Oasis reunion tour ticket prices
Ticketmaster's dynamic pricing system caused ticket prices for Oasis's reunion tour to surge from around £135 to over £488 for some fans. Fans reported paying double or triple the face value for standing tickets in Dublin and Cardiff. The system, which adjusts prices based on demand, has been criticised by fans on social media. Ticketmaster has not commented on the incident.
Company
Ticketmaster
9 source articlesnme.com

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

Facebook AI bans female photography charity Hundred Heroines due to word confusion

FTC settles with DoNotPay over deceptive AI lawyer claims

AI news presenters James and Rose terminated after two months at Hawaii newspaper

AskVet's AI chatbot VERA recommends euthanasia for woman's dog

Spinvox accused of using human transcribers for voice messages

Edinburgh Airport AI trial gives passengers different parking prices

Facebook failed to block thousands of political ads with false information

CDT study finds AI chatbots give inaccurate voting information to disabled voters

Elon Musk posts AI-generated image of Kamala Harris with false communist claim

Meta admits scraping Australian users' public data for AI training without opt-out

Mumsnet takes legal action against OpenAI for scraping content

Israeli settler group posts AI video of Al-Aqsa burning

FBI charges man for creating AI chatbots to harass university professor

Trump supporters share AI images of Trump saving cats and ducks

Amazon’s Alexa AI upgrade favored Harris over Trump in answer

Kevin Roose manipulates AI chatbots with invisible text on website

India-aligned network used AI fake accounts to target Pakistan, NewsGuard finds

Telegram deepfake porn crisis engulfs hundreds of South Korean schools

Steak 'n Shake sued over facial recognition kiosks under BIPA

Dutch DPA fines Clearview AI €30.5 million for GDPR violations

Microsoft Bing Copilot falsely accuses German journalist of crimes

BREIN takes down AI dataset for copyright infringement

Vanderbilt, Northwestern and University of Texas stop using Turnitin AI detector over false cheating accusations

Ticketmaster dynamic pricing inflates Oasis reunion tour ticket prices