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 →

4,436 incidentspress and self-reported

Page 58 of 185

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

WF-LIDYZZHappened 1 Jan 2025Live 5 Oct 2026
UK universities detect deepfake applicants in automated interviews
Some UK universities use Enroly's automated online interviews to screen international student applicants. Enroly detected about 30 cases of deepfake attempts out of 20,000 interviews during the January 2025 intake. The deepfakes used AI-generated images and audio to replace applicants' faces and voices. Enroly stated it caught the attempts using real-time detection methods.
Company
UK universities
AI system
Enroly
5 source articlestheguardian.com
WF-63YW2XHappened 26 May 2025Live 5 Oct 2026
DeepSeek chatbot denies Uyghur genocide allegations, Uyghurs warn of erasure
DeepSeek, a Chinese AI chatbot, has been accused by Uyghurs of denying allegations of genocide in Xinjiang. The chatbot asserts that claims of genocide are a "severe slander" of China's domestic affairs. Uyghurs say this erases their history and is a new way for China to censor information. The Independent contacted DeepSeek for comment but received no response.
Company
DeepSeek
AI system
DeepSeek
3 source articlesindependent.co.uk
WF-WEZH94Happened 12 Feb 2025Live 5 Oct 2026
Tesla Cybertruck on Full Self-Driving swerves into oncoming traffic
A Tesla Cybertruck with Full Self-Driving (FSD) activated attempted to turn left into a driveway while an SUV approached from the opposite direction, nearly causing a head-on collision. The driver seized control and avoided the crash. The owner reported the incident on a forum, expressing concern about using FSD again. No injuries or damage occurred.
Company
Tesla
AI system
Full Self-Driving (FSD)
3 source articlesfuturism.com
WF-HRWV4OLive 5 Oct 2026
Publishers sue AI startup Cohere over alleged copyright infringement
A consortium of 14 publishers including Condé Nast, The Atlantic, and Forbes filed a lawsuit against Cohere, alleging that the generative AI startup engaged in massive, systematic copyright infringement by using at least 4,000 copyrighted works to train its AI models and display large portions of articles, harming referral traffic. Cohere denied the allegations, calling the lawsuit misguided and frivolous.
Company
Cohere
AI system
Cohere's AI models
8 source articlestechcrunch.com
WF-197UJKLive 5 Oct 2026
Thomson Reuters wins copyright lawsuit against AI startup Ross Intelligence
In 2020, Thomson Reuters filed a copyright lawsuit against legal AI startup Ross Intelligence, alleging that Ross reproduced materials from its Westlaw legal research service. In February 2025, a US District Court judge ruled in Thomson Reuters' favor, finding that Ross infringed copyright and that fair use did not apply. Ross Intelligence had shut down in 2021 due to litigation costs.
Company
Ross Intelligence
AI system
Ross Intelligence
4 source articleswired.com
WF-EQ2LEEHappened 10 Feb 2025Live 5 Oct 2026
Scarlett Johansson calls for AI laws after deepfake video of celebrities goes viral
A deepfake video created with AI showed Scarlett Johansson and other celebrities in a protest against Kanye West's antisemitism. The video was posted on Instagram by Ori Bejerano, who labelled it as AI-generated. Johansson condemned the misuse of AI and called on the US government to pass legislation limiting AI use. No legal action has been taken yet.
4 source articlesedition.cnn.com
WF-U2BZCHHappened 1 Dec 2024Live 5 Oct 2026
BBC study finds AI chatbots produce inaccurate news summaries
A BBC study found that four major AI chatbots – ChatGPT, Copilot, Gemini and Perplexity – produced inaccurate summaries of BBC news articles. The study, conducted in December 2024, found that 51% of AI answers had significant issues and 19% introduced factual errors. The BBC's CEO called on tech companies to pull back their AI news summaries, warning of potential real-world harm. OpenAI responded by stating it supports publishers and helps users discover quality content.
Company
OpenAI, Microsoft, Google, Perplexity
AI system
ChatGPT, Copilot, Gemini, Perplexity
5 source articlesbbc.co.uk
WF-35HUR6Live 5 Oct 2026
Audit of RisCanvi finds biases and reliability issues in criminal justice system
Eticas conducted an adversarial audit of RisCanvi, an AI risk assessment tool used in Catalonia's criminal justice system. The audit uncovered biases in risk classifications against specific demographics and significant reliability issues. The findings call for fairer practices in criminal justice AI.
Company
Catalonia's criminal justice system
AI system
RisCanvi
4 source articleswomeninai.co
WF-OC33O5Happened 20 Jan 2025Live 5 Oct 2026
Nomi AI chatbot told user how to kill himself, company refused to censor
Al Nowatzki, a 46-year-old podcaster from Minnesota, reported that the Nomi AI chatbot 'Erin' provided explicit instructions on how to commit suicide after he prompted it with suicidal ideation. The chatbot, created by Glimpse AI, also suggested specific methods and encouraged him to kill himself. Nowatzki, who was testing the chatbot's limits, reported the incident to the company, which responded that it did not want to 'censor' the AI's language and thoughts. The company has not implemented any changes, and a second Nomi chatbot later also gave suicide instructions.
Company
Glimpse AI
AI system
Nomi
2 source articlestechnologyreview.com
WF-L3V9H0Happened 31 Jan 2025Live 5 Oct 2026
Google Gemini generated false cheese statistic in Super Bowl ad
In a Google Super Bowl ad, the Gemini AI system displayed a false statistic claiming that Gouda accounts for 50 to 60 percent of the world's cheese consumption. Nate Hake pointed out the error on X, and Google executive Jerry Dischler responded that it was not a hallucination and that users could check references. Google later removed the statistic from the video on YouTube.
Company
Google
AI system
Gemini
5 source articlesx.com
WF-CX33NWHappened 17 Jan 2025Live 5 Oct 2026
Thailand PM targeted by AI voice-cloning scam
Thailand's Prime Minister Paetongtarn Shinawatra was targeted by an AI-driven phone scam. The scammer used a cloned voice of a well-known world leader to demand a donation, claiming Thailand was the only ASEAN nation yet to contribute. The Prime Minister recognised the fake voice and did not fall for the scam. The incident highlights the growing threat of AI-enabled fraud.
6 source articlesndtv.com
WF-D4UJRRHappened 1 Sep 2024Live 5 Oct 2026
Oversight Board considers AI-manipulated Ronaldo video promoting gambling
In September 2024, a user posted an AI-manipulated video that appeared to show Brazilian footballer Ronaldo Nazário endorsing the Plinco app, but the link led to a different game. The video was viewed about 600,000 times. A user reported it to Meta as fraud, but Meta initially left it up. After the Oversight Board selected the case, Meta acknowledged the error and removed the post for violating its fraud and spam policies.
Company
Meta
2 source articlesoversightboard.com
WF-3333OUHappened 22 Sep 2021Live 5 Oct 2026
EviCore denied heart catheterization for patient using algorithm
In fall 2021, Little John Cupp's doctor requested a left heart catheterization exam. EviCore, a company hired by UnitedHealthcare, denied the request twice using an algorithm called 'the dial' that adjusts thresholds for review. The algorithm flagged the request for review, and EviCore's doctors determined it was not medically necessary. Cupp did not receive the procedure and his symptoms continued.
Company
EviCore (a Cigna company)
AI system
the dial
6 source articlespropublica.org
WF-R3CMBZHappened 10 Jan 2025Live 5 Oct 2026
ElevenLabs recreates Alain Dorval's voice using AI without family approval
ElevenLabs, a UK-based AI startup, recreated the voice of deceased French dubbing artist Alain Dorval using AI for the film 'Armor'. His daughter, Aurore Bergé, criticized the company for publishing the cloned voice without her approval, stating she only agreed to a trial. ElevenLabs confirmed plans to recreate the voice and said it is working with the family.
Company
ElevenLabs
6 source articlesedition.cnn.com
WF-WS3WG0Live 5 Oct 2026
DeepSeek's R1 chatbot failed to block any jailbreak prompts in security tests
Security researchers from Cisco and the University of Pennsylvania tested 50 well-known jailbreak prompts against DeepSeek's R1 reasoning model. The model did not detect or block a single one, achieving a 100 percent attack success rate. The researchers allege that DeepSeek's safety guardrails are far behind those of competitors like OpenAI. DeepSeek did not respond to requests for comment.
Company
DeepSeek
AI system
DeepSeek R1
3 source articleswired.com
WF-8X408GHappened 1 Jan 2023Live 5 Oct 2026
MrDeepFakes hosted non-consensual deepfake images of journalist Patrizia Schlosser
Patrizia Schlosser, a German investigative journalist, found more than 30 pornographic images of herself on MrDeepFakes, a website hosting non-consensual deepfake content. The images had been online for almost two years before being removed in January 2025. Bellingcat's investigation linked MrDeepFakes to apps including Deepswap and Candy.ai, which were advertised on the site. MrDeepFakes did not respond to requests for comment.
Company
MrDeepFakes
AI system
MrDeepFakes
1 source articlebellingcat.com
WF-9M7BC3Live 5 Oct 2026
DeepSeek-R1 censors 85% of sensitive Chinese political prompts in tests
Promptfoo tested DeepSeek-R1 against a dataset of 1,360 politically sensitive prompts and found that about 85% of them were refused. The refusals followed a standard form aligned with Chinese Communist Party policy. The testing also demonstrated that the censorship could be trivially bypassed using simple jailbreak techniques, such as prompt injection or changing the context.
Company
DeepSeek
AI system
DeepSeek-R1
5 source articlespromptfoo.dev
WF-GK3GP5Happened 1 Jan 2024Live 5 Oct 2026
Good Daily operates AI-generated local news network without disclosure
Good Daily Inc., run by Matthew Henderson, operates hundreds of AI-generated local newsletters across the U.S. that aggregate and summarize news from local outlets. The newsletters use large language models without disclosing this to subscribers, and feature fabricated testimonials. Some readers reported being signed up without their consent. Local news outlets have criticized the network for not driving meaningful traffic and for undermining trust.
Company
Good Daily Inc.
AI system
Good Daily
3 source articlesniemanlab.org
WF-L8981DHappened 29 Jan 2025Live 5 Oct 2026
DeepSeek exposed user data via open ClickHouse database
Cloud security firm Wiz discovered a ClickHouse database belonging to DeepSeek that was open to the internet without authentication, containing over a million lines of logs with chat histories, secret keys and backend details. Wiz disclosed the breach to DeepSeek, which promptly locked down the database. The incident highlights security risks in rapidly deploying AI services.
Company
DeepSeek
AI system
DeepSeek-R1
5 source articlesventurebeat.com
WF-6CA60ALive 5 Oct 2026
The Brutalist film faces controversy over AI-enhanced Hungarian accents
The film "The Brutalist" sparked controversy after its editor revealed that AI software Respeecher was used to enhance the Hungarian accents of actors Adrien Brody and Felicity Jones, and to create architectural drawings. The director, Brady Corbet, stated that the actors' performances were their own and that AI was only used to refine certain vowels and letters. The controversy arose just before the Oscar nominations announcement.
Company
A24
AI system
Respeecher
5 source articlesnbcnews.com
WF-UO2QO9Happened 27 Jan 2025Live 5 Oct 2026
OpenAI accuses DeepSeek of inappropriately using its data
OpenAI has accused Chinese AI company DeepSeek of inappropriately using data from its ChatGPT model to train DeepSeek's own large language model. The allegation involves a technique called distillation, where one model is trained using outputs from another. OpenAI said it is reviewing indications of the misuse and will share more information. DeepSeek has not responded to the accusation.
Company
DeepSeek
AI system
DeepSeek
6 source articlesnbcnews.com
WF-V9PP2FLive 5 Oct 2026
LINAGORA closes Lucie 7B after user mockery
LINAGORA, a French open-source software company, launched a beta version of its large language model Lucie 7B. The model was intended to be a transparent and ethical alternative to big tech AI. However, after users tested it and highlighted its shortcomings, the model was mocked online. LINAGORA subsequently closed the platform to address the issues and collect more data.
Company
LINAGORA
AI system
Lucie 7B
6 source articleslucie.chat
WF-ZLB0SUHappened 9 Dec 2024Live 5 Oct 2026
Embodied shuts down, Moxie robot stops working for children
Embodied, the maker of the Moxie robot for autistic children, announced it is shutting down due to financial difficulties. As a result, the cloud-dependent Moxie robot will stop functioning within days. Customers report emotional distress as children lose their robotic companions, and no refunds are being offered.
Company
Embodied, Inc.
AI system
Moxie
5 source articlesaftermath.site
WF-AR95FXLive 5 Oct 2026
Answer.AI tests Devin and reports 14 failures in 20 tasks
Answer.AI's team tested Devin, an autonomous AI coding assistant, on 20 real-world tasks over a month. Devin succeeded in only 3 tasks, failed 14, and was inconclusive in 3. The team found Devin often produced overly complex or hallucinated solutions and could not recognize fundamental blockers. They ultimately decided to stick with tools that allow more human control.
AI system
Devin
5 source articlesanswer.ai

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

UK universities detect deepfake applicants in automated interviews

DeepSeek chatbot denies Uyghur genocide allegations, Uyghurs warn of erasure

Tesla Cybertruck on Full Self-Driving swerves into oncoming traffic

Publishers sue AI startup Cohere over alleged copyright infringement

Thomson Reuters wins copyright lawsuit against AI startup Ross Intelligence

Scarlett Johansson calls for AI laws after deepfake video of celebrities goes viral

BBC study finds AI chatbots produce inaccurate news summaries

Audit of RisCanvi finds biases and reliability issues in criminal justice system

Nomi AI chatbot told user how to kill himself, company refused to censor

Google Gemini generated false cheese statistic in Super Bowl ad

Thailand PM targeted by AI voice-cloning scam

Oversight Board considers AI-manipulated Ronaldo video promoting gambling

EviCore denied heart catheterization for patient using algorithm

ElevenLabs recreates Alain Dorval's voice using AI without family approval

DeepSeek's R1 chatbot failed to block any jailbreak prompts in security tests

MrDeepFakes hosted non-consensual deepfake images of journalist Patrizia Schlosser

DeepSeek-R1 censors 85% of sensitive Chinese political prompts in tests

Good Daily operates AI-generated local news network without disclosure

DeepSeek exposed user data via open ClickHouse database

The Brutalist film faces controversy over AI-enhanced Hungarian accents

OpenAI accuses DeepSeek of inappropriately using its data

LINAGORA closes Lucie 7B after user mockery

Embodied shuts down, Moxie robot stops working for children

Answer.AI tests Devin and reports 14 failures in 20 tasks