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 →

4,738 incidentspress and self-reported

Page 121 of 198

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

WF-7J38Y4Happened 1 Aug 2018Live 5 Oct 2026
Study finds Amazon Rekognition has higher error rates for darker-skinned women
A study by Raji and Buolamwini found that Amazon's Rekognition facial analysis tool had a 31% error rate for classifying the gender of darker-skinned women, compared to 0% for lighter-skinned men. Amazon officials Matthew Wood and Michael Punke wrote blog posts disputing the findings, calling them misleading. The researchers called on Amazon to stop selling Rekognition to law enforcement.
Company
Amazon
AI system
Amazon Rekognition
10 source articlesmedium.com
WF-OC8AJTHappened 1 Dec 2012Live 5 Oct 2026
Surgeon sues Google over 'bankrupt' autocomplete suggestion
Dr Guy Hingston, a Port Macquarie surgeon, sued Google for defamation over an autocomplete suggestion that when users typed his name, "Guy Hingston bankrupt" appeared. He alleges the suggestion is false and has cost him patients and financiers. Google declined to comment. The lawsuit was filed in US district court in California.
Company
Google
AI system
Google Autocomplete
10 source articlessmh.com.au
WF-RKD0W6Happened 1 Mar 2017Live 5 Oct 2026
Uber suspends autonomous tests after Arizona crash
Uber suspended its self-driving car tests in Arizona and paused operations in Pittsburgh after one of its autonomous Volvo SUVs was involved in a high-impact crash in Tempe, Arizona. The crash occurred when another vehicle failed to yield, causing the Uber car to flip on its side. No injuries were reported, and the Uber vehicle was not at fault. Uber is investigating the incident.
Company
Uber Technologies Inc.
8 source articlesindustryweek.com
WF-C2R5FHLive 5 Oct 2026
Geico and other insurers charge higher premiums in minority neighborhoods than in white areas with similar risk
ProPublica's analysis of auto insurance premiums in California, Illinois, Texas and Missouri found that major insurers such as Geico, Allstate and Liberty Mutual charge minority neighborhoods up to 30% more than white areas with similar accident costs. The disparity is alleged to result from pricing algorithms that may inadvertently use zip code as a proxy for race. Otis Nash, a 26-year-old Chicagoan, pays $190.69 per month for Geico insurance while a white driver in a safer neighborhood pays $54.67. The insurers deny discrimination, stating they do not collect race data.
Company
Geico
5 source articlespropublica.org
WF-6M8JGSHappened 5 Oct 2017Live 5 Oct 2026
Mattel cancels Aristotle AI babysitter after privacy complaints
Mattel cancelled its Aristotle AI assistant for babies after privacy advocates and child psychologists raised concerns. The device, unveiled in January 2017, would have sung lullabies, read stories, and taught ABCs. A petition against it gained over 17,000 signatures. Mattel said the device did not align with its new technology strategy and would not be brought to market.
Company
Mattel
AI system
Aristotle
10 source articlestheverge.com
WF-IIYZYMHappened 1 Jan 2016Live 5 Oct 2026
Ajin USA pleads guilty after worker crushed by robot in Alabama
In 2016, 20-year-old Regina Allen Elsea was killed at the Ajin USA auto parts plant in Cusseta, Alabama, when a robotic machine abruptly restarted and crushed her while she was clearing a sensor fault. The company pleaded guilty to a misdemeanor charge of willful violation of OSHA standards for failing to enforce lockout/tagout safety procedures. Ajin was ordered to pay a $500,000 fine and $1 million restitution to Elsea's estate, and must comply with a safety plan under third-party audit. A separate civil lawsuit by her family is ongoing.
Company
Ajin USA
10 source articlesapnews.com
WF-ZNK9ANLive 5 Oct 2026
Stanford researcher builds AI to detect sexual orientation from faces
Stanford researcher Michal Kosinski and Yilun Wang built an AI system using VGG-Face that claims to detect sexual orientation from facial images. The study, which used 15,000 photos from a dating website, has been criticized by AI researchers and LGBTQ groups for potential privacy violations and flawed methodology. Kosinski says he built the system to highlight the dangers of privacy-infringing AI.
Company
Stanford University
AI system
VGG-Face
10 source articlesqz.com
WF-FB00CQHappened 1 Dec 2017Live 5 Oct 2026
Australian education ministers scrap NAPLAN robot marking
The Education Council decided not to use automated essay scoring for NAPLAN writing tests, dropping a proposal by ACARA to have computers mark English tasks. Teachers' unions had campaigned against the plan, and an academic warned that the software would reward verbose gibberish and could not assess creativity. The decision was made in December 2017 and announced in January 2018.
Company
Australian Curriculum, Assessment and Reporting Authority (ACARA)
10 source articlesabc.net.au
WF-L5NTEZHappened 27 Oct 2014Live 5 Oct 2026
Teachers find no value in SAS EVAAS system in Southwest School District
A study examined the SAS Education Value-Added Assessment System (EVAAS®) used by the Southwest School District (SSD) to evaluate teacher effectiveness for high-stakes consequences. Teachers reported that the system produced inconsistent results, was biased by student factors, and reduced morale and collaboration. The study found that teachers did not use the data formatively as intended, and unintended consequences included heightened pressure and teaching to the test.
Company
Southwest School District (SSD)
AI system
SAS Education Value-Added Assessment System (EVAAS®)
10 source articlesepaa.asu.edu
WF-0JU6DRHappened 20 Mar 2017Live 5 Oct 2026
Chinese park uses facial recognition to limit toilet paper theft
The Temple of Heaven Park in Beijing installed facial recognition software on toilet paper dispensers to prevent visitors from taking excessive amounts. The system allows one serving per person every nine minutes. Visitors expressed privacy concerns but acknowledged the need to curb theft by older people and migrant workers.
Company
Temple of Heaven Park
10 source articlesedition.cnn.com
WF-DJVN67Live 5 Oct 2026
Australian government settles Robodebt class action for $1.7 billion
The Robodebt system, an automated debt recovery system used by Centrelink, calculated welfare debts using averaged income data from the ATO, leading to invalid debts. A class action was filed in November 2019 and settled in November 2020. The settlement provided refunds of over $751 million in invalid debts collected and $112 million in compensation to approximately 400,000 group members. The Australian government also dropped claims for over $1 billion in invalid debts.
Company
Australian Government (Centrelink)
AI system
Robodebt
10 source articlesgordonlegal.com.au
WF-4UX2JALive 5 Oct 2026
YouTube's algorithm systematically exposes children to disturbing automated content
YouTube's recommendation algorithm and automated content generation systems have been producing and promoting disturbing videos targeting children. These videos, often generated by bots using keyword stuffing and stock animations, include frightening and abusive content. The article alleges that the platform's automated systems, without human oversight, systematically traumatise children at scale.
Company
YouTube
AI system
YouTube recommendation algorithm
10 source articlesmedium.com
WF-SIARFNHappened 1 Jan 2012Live 5 Oct 2026
China police collect voice biometrics with iFlytek for mass surveillance
The Chinese government, collaborating with AI company iFlytek, has been collecting voice pattern samples to build a national voice biometric database. Police use an Automatic Speaker Recognition system to automatically identify targeted voices in phone conversations and alert authorities. Human Rights Watch reports that the collection occurs with little transparency, and individuals may not know their voice patterns have been taken or be able to challenge the surveillance. iFlytek has not responded to Human Rights Watch's inquiries about the program.
Company
Ministry of Public Security
AI system
Automatic Speaker Recognition (ASR) system
10 source articleshrw.org
WF-PZRPZRHappened 1 Jan 2017Live 5 Oct 2026
Royal Free London publishes audit into Streams app data processing
The Royal Free London NHS Foundation Trust published an audit into its use of the Streams app, following an investigation by the Information Commissioner's Office (ICO) in July 2017. The Streams app alerts clinicians to patients at risk of acute kidney injury. The audit, conducted by Linklaters, concluded that the trust's use of Streams was lawful and complied with data protection laws, although areas for improvement were identified. The ICO later recognised that the trust had completed all required actions.
Company
Royal Free London NHS Foundation Trust
AI system
Streams
10 source articlesroyalfree.nhs.uk
WF-0UD8YWLive 5 Oct 2026
TrueCompanion's Roxxxy doll Frigid Farrah simulates rape
The Roxxxy TrueCompanion sex robot includes a personality setting called Frigid Farrah, which is programmed to resist the user's advances, simulating rape. The doll is sold online for $9,995. The article criticizes the product as depraved and calls for policy consideration.
Company
TrueCompanion
AI system
Roxxxy TrueCompanion with Frigid Farrah personality
10 source articlesmarieclaire.com.au
WF-9LDJW1Live 5 Oct 2026
Man faces 10 years in jail for AI crocodile image
A man faces a 10-year jail sentence for an AI-generated crocodile image, as reported by Popular Science.
7 source articlespopsci.com
WF-31WCUHLive 5 Oct 2026
New York teacher challenges state evaluation system over flawed rating
Sheri Lederman, a 4th-grade teacher in the Great Neck district, filed a lawsuit against the New York state education department after being rated 'ineffective' on the portion of her evaluation tied to students' test performance, a year after receiving a 'highly effective' score. The lawsuit alleges that the teacher-evaluation system is statistically flawed and lacks a proper appeals process. The education department declined to comment on the litigation.
Company
New York state education department
10 source articlesedweek.org
WF-78BIQPHappened 3 Jun 2017Live 5 Oct 2026
Uber's pricing algorithm surged prices during London Bridge attack
On June 3, 2017, during a terrorist attack on London Bridge, Uber's dynamic pricing algorithm increased ride prices by over 200% for 43 minutes. Customers attempting to flee the area faced surge pricing. The algorithm automatically adjusted prices based on supply and demand without accounting for the emergency.
Company
Uber
7 source articleshbr.org
WF-2865P3Happened 1 Aug 2012Live 5 Oct 2026
Knight Capital fined $12 million for 2012 trading glitch that caused $460 million loss
On August 1, 2012, Knight Capital Americas LLC's automated equity router malfunctioned due to a code deployment error, sending over 4 million erroneous orders into the market within 45 minutes. The firm suffered a loss of more than $460 million. The SEC charged Knight Capital with violating the market access rule and Regulation SHO, and the firm agreed to pay a $12 million penalty and retain an independent consultant.
Company
Knight Capital Americas LLC
AI system
Knight Capital's automated equity router
10 source articlessec.gov
WF-0K47RDLive 5 Oct 2026
Idaho Medicaid used secret formula to cut assistance for disabled adults
The Idaho Department of Health and Welfare used a secret formula to drastically cut Medicaid assistance for adults with developmental disabilities, claiming the formula was a trade secret. The ACLU of Idaho filed a class action lawsuit, K.W. v. Armstrong, in 2012, alleging the cuts violated due process. A federal court struck down the formula in 2016 and ordered a new system, but as of 2024, a fair system has not yet been implemented.
Company
Idaho Department of Health and Welfare
10 source articlesacluidaho.org
WF-54EB8ULive 5 Oct 2026
Target's pregnancy prediction algorithm reveals teen's pregnancy to father
Target used a data mining algorithm to predict when customers were pregnant based on their purchases. The algorithm sent baby product coupons to a high school student, leading her father to complain to the store. The father later learned his daughter was indeed pregnant. Target subsequently altered the coupons to disguise the data mining.
Company
Target
AI system
Target pregnancy prediction algorithm
9 source articlesabc.net.au
WF-9PIBK7Happened 1 Dec 2012Live 5 Oct 2026
Google found liable for defamatory search results in Australia
Google's search engine displayed images of Milorad Trkulja alongside convicted criminals and suggested defamatory autocomplete predictions. Trkulja sued Google for defamation. The High Court of Australia ruled that Google could be considered a publisher of the search results and that the material was capable of being defamatory, allowing the case to proceed.
Company
Google LLC
AI system
Google Search
10 source articlesglobalfreedomofexpression.columbia.edu
WF-VX03JZHappened 1 Jan 2009Live 5 Oct 2026
Google Search results for 'black girls' and 'three black teenagers' showed pornographic and mugshot images
Google Search returned pornographic results for the query 'black girls' and mugshots for 'three black teenagers', reflecting racial and gender stereotypes. The author, Safiya Noble, documented these results from 2009 to 2016. Google initially denied responsibility but later tweaked its algorithm to change the results. The incident highlights systemic bias in search algorithms.
Company
Google
AI system
Google Search
8 source articlestime.com
WF-MDQFY6Happened 1 Jan 2013Live 5 Oct 2026
Bettina Wulff settles with Google over Autocomplete name associations
Bettina Wulff, wife of former German president Christian Wulff, sued Google over its Autocomplete feature that combined her name with terms from the red-light district. The search engine automatically generated such suggestions. Google and Wulff reached an out-of-court settlement, and Google independently changed its guidelines. The Federal Court of Justice had ruled in 2013 that Google must delete search suggestions that directly violate personality rights.
Company
Google
AI system
Autocomplete
10 source articlessueddeutsche.de

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

Study finds Amazon Rekognition has higher error rates for darker-skinned women

Surgeon sues Google over 'bankrupt' autocomplete suggestion

Uber suspends autonomous tests after Arizona crash

Geico and other insurers charge higher premiums in minority neighborhoods than in white areas with similar risk

Mattel cancels Aristotle AI babysitter after privacy complaints

Ajin USA pleads guilty after worker crushed by robot in Alabama

Stanford researcher builds AI to detect sexual orientation from faces

Australian education ministers scrap NAPLAN robot marking

Teachers find no value in SAS EVAAS system in Southwest School District

Chinese park uses facial recognition to limit toilet paper theft

Australian government settles Robodebt class action for $1.7 billion

YouTube's algorithm systematically exposes children to disturbing automated content

China police collect voice biometrics with iFlytek for mass surveillance

Royal Free London publishes audit into Streams app data processing

TrueCompanion's Roxxxy doll Frigid Farrah simulates rape

Man faces 10 years in jail for AI crocodile image

New York teacher challenges state evaluation system over flawed rating

Uber's pricing algorithm surged prices during London Bridge attack

Knight Capital fined $12 million for 2012 trading glitch that caused $460 million loss

Idaho Medicaid used secret formula to cut assistance for disabled adults

Target's pregnancy prediction algorithm reveals teen's pregnancy to father

Google found liable for defamatory search results in Australia

Google Search results for 'black girls' and 'three black teenagers' showed pornographic and mugshot images

Bettina Wulff settles with Google over Autocomplete name associations