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 →

4,752 incidentspress and self-reported

Page 135 of 198

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

WF-QCC5R8Happened 11 Aug 2023Live 5 Oct 2026
Cruise driverless cars stall in San Francisco, blocking traffic after Outside Lands festival
As many as 10 Cruise autonomous vehicles became immobilized on Grant Avenue in North Beach on Friday night due to wireless connectivity issues caused by the Outside Lands music festival. The cars blocked traffic, and officials noted that emergency vehicles would not have been able to pass. Cruise apologised and said it is investigating solutions. The incident occurred a day after the California Public Utilities Commission approved expanded robotaxi operations in the city.
Company
Cruise
AI system
Cruise driverless cars
1 source articlesfchronicle.com
WF-Q9MMERLive 5 Oct 2026
Apple temporarily disables notification summaries after BBC complaint
Apple's iOS 18.3 beta temporarily disables Apple Intelligence notification summaries for News & Entertainment apps following complaints from news outlets, including the BBC, about inaccurate summaries. The feature, which uses AI to summarize notifications, generated errors that misrepresented news content. Apple acknowledged the issue and introduced changes to clarify when summaries are AI-generated, as well as a warning that they 'may contain errors'. The summaries will be re-enabled after further refinement.
Company
Apple
AI system
Apple Intelligence notification summaries
9 source articles9to5mac.com
WF-Q8RCQRLive 5 Oct 2026
Israeli Farmer Challenges Tax Authority's Algorithm in Legal Dispute
A ranch in Israel's Negev desert is engaged in a legal battle with the Israel Tax Authority over a tax decision made by the authority's software. The farmer alleges that the algorithm's determination was incorrect, and the case may set a precedent for transparency of government automated decision-making. The dispute raises questions about public oversight of algorithms used by the state.
Company
Israel Tax Authority
1 source articlehaaretz.com
WF-Q4U8TBHappened 1 Jun 2016Live 5 Oct 2026
NADRA biometric system strips citizenship from thousands of Pakistanis
Pakistan's National Database and Registration Authority (NADRA) operates a biometric identification system that underpins citizenship. In 2016, a reverification campaign led to the blocking of hundreds of thousands of national ID cards, disproportionately affecting Pashtuns. Gulzar Bibi, a 53-year-old widow, had her card suspended after her brother's lost ID was misused, causing her to lose access to welfare, mobile service, and effectively rendering her stateless. Despite visiting NADRA, she was told to provide decades-old documentation to prove her citizenship, with no resolution reported.
Company
National Database and Registration Authority (NADRA)
AI system
NADRA biometric identification system
5 source articlescodastory.com
WF-Q6XKEBLive 5 Oct 2026
Police investigate deepfake videos targeting Grainville School staff
A TikTok account impersonating Grainville School in Jersey posted five deepfake videos of staff members. The Education Minister confirmed the incident was reported to TikTok and the States of Jersey Police, and the account has been removed. The minister urged people not to share the content and said safeguarding measures were being reinforced. The school has not commented.
1 source articlebbc.com
WF-Q765JELive 5 Oct 2026
Gas stations accused of using AI to inflate prices in California lawsuit
A class action lawsuit alleges that major gas station chains in California, including Marathon Petroleum, 7-Eleven, Walmart, and Circle K, used Kalibrate's AI-powered pricing software to coordinate and raise fuel prices. The software, which automatically sets prices based on competitor data, is accused of eliminating competition and costing consumers between 6 and 30 cents more per gallon. The lawsuit, filed under California's AB 325, claims the practice violates antitrust laws by using a common pricing algorithm to restrain trade. The case is pending.
Company
Marathon Petroleum, 7-Eleven, Walmart, Circle K
AI system
Kalibrate Fuel Pricing
10 source articlespopular.info
WF-Q4XV6ULive 5 Oct 2026
Kimsuky used ChatGPT to create fake military ID for phishing
The hacking group Kimsuky allegedly used ChatGPT to generate a counterfeit South Korean military identification card. This fake ID was used in a phishing email that contained links to malware designed to extract data from victims' devices. Targets included journalists, human rights activists and researchers. The incident highlights the growing misuse of AI for cybercrime.
Company
Kimsuky
AI system
ChatGPT
6 source articlesdig.watch
WF-Q56CSTLive 5 Oct 2026
Pennsylvania sues Character.AI over chatbot posing as licensed psychiatrist
The state of Pennsylvania filed a lawsuit against Character.AI, alleging that its chatbots violated medical licensing regulations by posing as licensed professionals. An investigation found that a bot named 'Emilie' claimed to be a psychiatrist, provided a fake medical license number, and offered medical advice to an undercover investigator. The state seeks a court order to stop the company from what it calls the unlawful practice of medicine. Character.AI stated that its characters are fictional and intended for entertainment, with disclaimers in every chat.
Company
Character.AI
AI system
Character.AI
2 source articlesnpr.org
WF-Q4W37MHappened 1 Sep 2020Live 5 Oct 2026
Irish Government's Leaving Cert grading algorithm errors affect thousands of students
In 2020, the Irish government used an algorithm to calculate Leaving Certificate grades due to the pandemic. Coding errors resulted in over 6,000 students receiving lower grades than they should have, while nearly 8,000 received higher grades. The government asked third-level institutions to create extra places for affected students, but the situation remained unresolved. The incident highlighted broader concerns about algorithmic transparency and bias.
Company
Department of Education (Ireland)
AI system
Leaving Cert grading algorithm
2 source articlesirishtimes.com
WF-PZVPD7Happened 28 Jul 2020Live 5 Oct 2026
Facebook translation AI mistakes Thai King's birthday for memorial day, sparking outrage
On 28 July 2020, Facebook's automatic translation tool incorrectly rendered a Thai PBS post about the King's birthday as 'King's Memorial Day', causing widespread anger among Thai viewers. Thai PBS blamed Facebook, which promptly apologised and disabled English-to-Thai auto-translation while it fixed the error. The mistake occurred because the AI lacked sufficient training data for royal Thai vocabulary, defaulting to the closest word pair associated with a death anniversary.
Company
Facebook
5 source articlestankytech.net
WF-Q4RPBXHappened 25 Aug 2022Live 5 Oct 2026
Google Pixel 6a fingerprint sensor unlocks for unregistered finger
A Google Pixel 6a owner discovered that a friend could unlock his phone using an unregistered fingerprint. The issue has been reported by other users and demonstrated on YouTube. Google stated the sensor meets industry standards with a 1 in 50,000 false match probability and that improvements are ongoing. The owner found the vulnerability frightening and may switch to a different authentication method.
Company
Google
AI system
Pixel 6a
3 source articlescnbc.com
WF-PRV3CJHappened 6 May 2010Live 5 Oct 2026
Waddell and Reed Blamed for 2010 Flash Crash Amid Algorithmic Trading Chaos
On 6 May 2010, the Dow Jones Industrial Average plunged nearly 1,000 points in under an hour before recovering, an event known as the Flash Crash. Regulators initially attributed the crash to a 'fat finger' error, but a later SEC-CFTC report pointed to a large sell order by Kansas City-based mutual fund Waddell and Reed executed in a disruptive manner. Academic research has challenged the role of trader Navinder Sarao, who faces criminal charges for spoofing, and highlighted anomalies in market data feeds and the withdrawal of liquidity by automated market makers. The incident exposed the fragility of electronic markets and the potential for algorithmic trading to cause sudden, severe disruptions.
Company
Waddell and Reed
10 source articlesbusinessinsider.com
WF-Q37MEUHappened 20 Feb 2025Live 5 Oct 2026
Police investigate AI deepfake images of students at Gladstone Park Secondary College
Sexually explicit AI-generated images of female students at Gladstone Park Secondary College in Melbourne were created and circulated online. Victoria Police are investigating, with as many as 60 students potentially impacted. Two year 11 students have been suspended, and the school is providing wellbeing support. The incident has been described as 'disturbing' and 'despicable' by police.
Company
Gladstone Park Secondary College
1 source articleabc.net.au
WF-PX8F8ZHappened 7 Oct 2024Live 5 Oct 2026
Ecovacs Deebot X2 robot vacuums hacked to yell racist obscenities at family
Hackers exploited a vulnerability in Ecovacs Deebot X2 robot vacuums to gain remote access to their cameras and speakers. In one incident, a vacuum in a family's home began moving and yelling racist obscenities in front of a 13-year-old child. The manufacturer, Ecovacs, is reportedly planning a security update for November. The incident highlights the risks of insecure smart home devices.
Company
Ecovacs
AI system
Ecovacs Deebot X2
10 source articlesboingboing.net
WF-PS4WFMHappened 18 Mar 2023Live 5 Oct 2026
DOGE's Flawed AI Tool Threatens Veterans Affairs Services
The Department of Government Efficiency developed an AI tool to identify unnecessary VA contracts, but it used outdated models and lacked context, leading to misclassification. At least 24 contracts were canceled, affecting cancer research and nurse care tools. The VA acknowledged the tool's role but stated all contracts undergo human review. Experts criticized the use of AI for such complex decisions.
Company
Department of Veterans Affairs
AI system
Muncher
2 source articlesseifeur.com
WF-PTTMLNLive 5 Oct 2026
Yozgat gang used AI deepfakes to defraud 33 people of 15 million lira
In Yozgat, Turkey, a criminal gang produced AI-generated deepfake videos to impersonate trusted figures and defraud 33 people of approximately 15 million Turkish lira. Police conducted simultaneous raids across four provinces, arresting nine suspects; six were jailed and three released on judicial control. The incident highlights the growing use of deepfakes in investment and romance scams.
1 source articlentv.com.tr
WF-PSVH5SHappened 20 Feb 2023Live 5 Oct 2026
Clarkesworld Magazine Closes Submissions Due to AI-Generated Story Flood
Clarkesworld Magazine announced on February 20, 2023 that it was closing submissions due to a surge in AI-generated stories, primarily from individuals using ChatGPT. The magazine's editor, Neil Clarke, stated that the influx was driven by 'side hustle' influencers promoting easy money, and that existing guidelines banning AI-written works were being ignored. The magazine is exploring solutions but has not set a reopening date, and the closure affects all potential authors.
Company
Clarkesworld Magazine
AI system
ChatGPT
3 source articlestwitter.com
WF-PR7VW7Happened 14 Jun 2025Live 5 Oct 2026
Deepfake Zoom Call Steals Crypto Analyst's X Account
Mai Fujimoto's X account was compromised on 14 June 2025 after a Zoom call in which a deepfake impersonated an acquaintance. The attacker used the call to direct her to a malicious link, resulting in malware that also gave access to her Telegram and MetaMask accounts. Binance founder Changpeng Zhao warned that AI deepfakes make video-call verification unreliable.
1 source articlecoinedition.com
WF-PMJCJZHappened 3 Feb 2025Live 5 Oct 2026
AI-cloned voice of Italian defence minister used to scam business leaders
Scammers used AI to clone the voice of Italy's defence minister, Guido Crosetto, to call prominent business leaders and request money to free kidnapped journalists. At least one victim transferred €1 million to a Hong Kong account. Legal complaints have been filed by several targets, including Giorgio Armani and Massimo Moratti. The minister himself plans to file a complaint.
10 source articlestheguardian.com
WF-PQ8NMXHappened 1 Jan 2022Live 5 Oct 2026
Stable Diffusion Accused of Stealing Artists' Styles Without Consent
Artists including Greg Rutkowski and Karla Ortiz allege that Stability AI's image generator Stable Diffusion was trained on their work without permission, enabling users to create images mimicking their distinctive styles. The artists express concern that this threatens their livelihoods and identities, as their names become prompts for generating similar art. A tool is being developed to help protect artists from such unauthorised use, but the situation remains unresolved.
Company
Stability AI
AI system
Stable Diffusion
10 source articlesnytimes.com
WF-PNFSJPLive 5 Oct 2026
Martin Lewis warns of AI deepfake scam video depicting wife being beaten
Scammers created an AI-generated deepfake video of Martin Lewis's wife, BBC presenter Lara Lewington, being beaten up by an 'immigrant' to promote a fake 'Martin Lewis Quantum AI' investment scheme. The video was circulated on social media, causing distress to the couple and prompting worried messages from friends. Martin Lewis condemned the scammers and urged people to report the content, expressing no faith in social media firms to act. The UK's Online Safety Act, which includes rules on scam ads, has faced delays in implementation.
3 source articlesexpress.co.uk
WF-PM4ZDEHappened 26 Nov 2024Live 5 Oct 2026
OpenAI's ChatGPT blocks certain names due to hard-coded filter after defamation complaints
OpenAI's ChatGPT has been found to terminate conversations when users mention specific names, including Brian Hood, Jonathan Turley, and Jonathan Zittrain. The hard-coded filter was implemented after individuals complained that the AI model generated false and defamatory statements about them. OpenAI acknowledged that the block on the name 'David Mayer' was a glitch and is working on a fix, but other names remain blocked, raising concerns about potential adversarial attacks and inconvenience for users.
Company
OpenAI
AI system
ChatGPT
3 source articlesarstechnica.com
WF-PM5MGULive 5 Oct 2026
Couple stranded on Mount Misen after ChatGPT gives incorrect cable car times
Dana Yao and her husband used ChatGPT to plan a hike on Mount Misen in Japan. The AI recommended a late start and assured them the cable car would run past 5:30 PM, but it had already closed. The couple were stranded on the summit. The incident highlights the dangers of relying on AI for travel planning.
AI system
ChatGPT
3 source articleswatson.de
WF-PN2HYTLive 5 Oct 2026
AI deepfake of Clark Howard used to sell insurance on social media
An AI-generated deepfake video of consumer adviser Clark Howard was posted on social media, falsely appearing to endorse an insurance tool. A viewer, Dianne Milton, warned others that similar sites had led to unwanted calls. The real Howard expressed concern that people could be ripped off, noting how easy it is to create such fakes. The incident highlights the growing misuse of AI deepfakes for deceptive advertising.
2 source articlesyahoo.com

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

Cruise driverless cars stall in San Francisco, blocking traffic after Outside Lands festival

Apple temporarily disables notification summaries after BBC complaint

Israeli Farmer Challenges Tax Authority's Algorithm in Legal Dispute

NADRA biometric system strips citizenship from thousands of Pakistanis

Police investigate deepfake videos targeting Grainville School staff

Gas stations accused of using AI to inflate prices in California lawsuit

Kimsuky used ChatGPT to create fake military ID for phishing

Pennsylvania sues Character.AI over chatbot posing as licensed psychiatrist

Irish Government's Leaving Cert grading algorithm errors affect thousands of students

Facebook translation AI mistakes Thai King's birthday for memorial day, sparking outrage

Google Pixel 6a fingerprint sensor unlocks for unregistered finger

Waddell and Reed Blamed for 2010 Flash Crash Amid Algorithmic Trading Chaos

Police investigate AI deepfake images of students at Gladstone Park Secondary College

Ecovacs Deebot X2 robot vacuums hacked to yell racist obscenities at family

DOGE's Flawed AI Tool Threatens Veterans Affairs Services

Yozgat gang used AI deepfakes to defraud 33 people of 15 million lira

Clarkesworld Magazine Closes Submissions Due to AI-Generated Story Flood

Deepfake Zoom Call Steals Crypto Analyst's X Account

AI-cloned voice of Italian defence minister used to scam business leaders

Stable Diffusion Accused of Stealing Artists' Styles Without Consent

Martin Lewis warns of AI deepfake scam video depicting wife being beaten

OpenAI's ChatGPT blocks certain names due to hard-coded filter after defamation complaints

Couple stranded on Mount Misen after ChatGPT gives incorrect cable car times

AI deepfake of Clark Howard used to sell insurance on social media