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 →

4,463 incidentspress and self-reported

Page 99 of 186

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

WF-CE1MQ4Happened 7 Jul 2016Live 5 Oct 2026
Dallas police used robot to kill gunman after standoff
In July 2016, the Dallas Police Department used a remote-controlled bomb disposal robot to deliver an explosive device to kill Micah Xavier Johnson, a gunman who had killed five police officers and was barricaded in a parking garage. Police Chief David Brown stated that the tactic was used as a last resort to avoid exposing officers to further danger. The incident raised ethical questions about the use of robots for lethal force by law enforcement.
Company
Dallas Police Department
8 source articleslatimes.com
WF-CGL1WPHappened 21 Jan 2021Live 5 Oct 2026
Cruzcampo deepfake ad of Lola Flores goes viral
Cruzcampo launched an advertising campaign using a deepfake video of the late singer Lola Flores, created with artificial intelligence. The video went viral shortly after release. The company explained the technical process and involved Flores' daughters in the project.
Company
Cruzcampo
10 source articleselpais.com
WF-1E03SDLive 5 Oct 2026
Kohler, BMW, MaxMara secretly collected customers' facial recognition data
During the 2021 CCTV 3·15 Gala, it was revealed that Kohler, BMW, and MaxMara stores had installed facial recognition cameras from vendors such as Wandianzhang and Youluoke. These cameras captured customers' facial data without their knowledge or consent, and the data was used to track customer visits and inform sales strategies. The systems were deployed in thousands of stores across China, collecting over 100 million facial records. The companies did not inform customers or obtain consent, violating Chinese privacy laws.
Company
Kohler (China) Investment Co., Ltd., ZhengTong Auto (BMW dealerships), MaxMara
AI system
Facial recognition cameras from Wandianzhang, Youluoke, Yaliang, Ruiwei
10 source articlesnews.cctv.com
WF-BWQHCVHappened 23 Feb 2021Live 5 Oct 2026
AnyVision patents drone facial recognition technology
AnyVision, an Israeli private surveillance company, filed a US patent for drone technology that uses facial recognition to identify people on the ground. The patent describes adaptive positioning of drones to find optimal angles for accurate facial recognition. The technology is not yet in production, but the company's CEO stated it will become a reality soon. Potential privacy concerns have been raised.
Company
AnyVision
AI system
Adaptive positioning of drones for enhanced facial recognition
9 source articlescyprus-mail.com
WF-84AKHBLive 5 Oct 2026
Seoul Metropolitan Government rolls out AI CCTV cameras to prevent suicides
The Seoul Metropolitan Government has rolled out AI-enabled CCTV cameras on Han River bridges to identify people at risk of suicide. The system uses deep learning to analyse behavioural patterns and alerts rescue teams. Experts and Privacy International have raised concerns that it is invasive and could be misused because it captures biometric data without explicit consent. The article also considers whether such technology could be implemented in India.
Company
Seoul Metropolitan Government
8 source articlesthequint.com
WF-J9BAYQHappened 29 Mar 2021Live 5 Oct 2026
Deepfake accounts pretending to be Amazon workers cause confusion on Twitter
In March 2021, ahead of a union vote at an Amazon warehouse in Bessemer, Alabama, Twitter accounts using deepfake profile pictures appeared, purporting to be Amazon employees. The accounts posted over-the-top defenses of Amazon's working practices, leading to public confusion about whether Amazon was behind an anti-union social media campaign. Amazon denied involvement, and the accounts were later suspended. The article notes that the accounts were likely parodies, not part of a corporate strategy.
10 source articlestechnologyreview.com
WF-FH1O73Happened 1 Jan 2017Live 5 Oct 2026
DoorDash changes tipping policy after backlash over worker tips
DoorDash, a food delivery company, changed its tipping policy after backlash. The company had been using customer tips to offset its base pay to delivery workers, meaning workers did not receive the full tips. After public outcry and promises from CEO Tony Xu, DoorDash announced a new model where workers will keep all tips, effective next month.
Company
DoorDash
10 source articlescnet.com
WF-979P50Live 5 Oct 2026
ByteDance whistleblower alleges Chinese government uses TikTok algorithms to censor content
A ByteDance whistleblower alleges that the company's algorithms automatically identify and delete content critical of the Chinese government, including posts in the Uighur language. The whistleblower claims that the Cyberspace Administration of China issues directives to ByteDance's Content Quality Center, which then tasks teams with applying censorship instructions to both ongoing and past content. The allegations suggest that ByteDance is suppressing minorities and masking human rights violations through its content moderation systems.
Company
ByteDance
AI system
Content moderation algorithms
8 source articlesfoxnews.com
WF-W8NDC1Happened 1 Jul 2015Live 5 Oct 2026
New York City's McKinsey-led jail violence program manipulated data, violence increased
New York City paid McKinsey & Company $27.5 million to reduce violence at Rikers Island jail complex. McKinsey designed a predictive algorithm called the Housing Unit Balancer and Restart housing units, but jail officials and McKinsey consultants stacked the units with compliant inmates to artificially lower violence numbers. Violence actually increased by nearly 50% during the project. The city eventually decided to close Rikers.
Company
New York City Department of Correction
AI system
Housing Unit Balancer (HUB)
10 source articlesmotherjones.com
WF-NVKO9DHappened 16 Dec 2020Live 5 Oct 2026
Facebook's policy change brought partisan content back to Georgia feeds
The Markup investigated Facebook's decision to lift its political ad ban in Georgia ahead of the Senate runoff elections. The study found that after December 16, partisan campaign content and misinformation displaced traditional news sources in users' feeds. Facebook acknowledged the change but did not reverse it.
Company
Facebook
AI system
Facebook News Feed
10 source articlesthemarkup.org
WF-N16FBJHappened 1 Oct 2022Live 5 Oct 2026
Israeli military uses Smash Hopper turret against Palestinian protesters
The Israeli military deployed Smash Hopper remote-controlled gun turrets in the West Bank. The turrets fire stun grenades, less-than-lethal bullets, and tear gas at Palestinian protesters without warning. The systems were observed at Al-Aroub refugee camp and Hebron in October 2022. The Smash Hopper is produced by Smart Shooter.
Company
Israeli military
AI system
Smash Hopper
10 source articlesmedium.com
WF-U3W24EHappened 18 Feb 2021Live 5 Oct 2026
Facebook's automated news ban blocks non-news pages in Australia
Facebook implemented an automated news ban in Australia in response to the government's news media code, blocking pages it classified as news. This incorrectly affected hundreds of non-news pages, including a funeral business and government health pages. Facebook acknowledged the error and said it would review and unblock pages, with an appeals process to be launched.
Company
Facebook
10 source articlestheguardian.com
WF-9HOEYDHappened 1 Dec 2020Live 5 Oct 2026
Facebook continued recommending political groups despite promise to stop
Facebook promised to stop recommending political groups to users in October 2020, but according to data from The Markup's Citizen Browser project, the platform continued to recommend such groups throughout December 2020 and January 2021. The recommendations disproportionately targeted Trump voters, and some groups contained conspiracy theories and calls to violence. Facebook acknowledged the issue and said it was investigating.
Company
Facebook
9 source articlesthemarkup.org
WF-BX55BUHappened 1 Mar 2022Live 5 Oct 2026
Serbia's Social Card law could harm marginalized groups
Serbia's Social Card law, which entered into force on 1 March 2022, establishes a centralized government database to assess eligibility for social security support. Amnesty International and seven other rights organizations submitted a legal opinion to the Constitutional Court, alleging that the automated system is an intrusive surveillance system that could discriminate against Roma communities and people with disabilities. The system processes 130 categories of personal data and lacks transparency, potentially leading to errors and denial of benefits.
Company
Serbian government
AI system
Social Card system
10 source articlesamnesty.org
WF-3281BZHappened 1 Feb 2021Live 5 Oct 2026
Facebook's ad system erroneously blocks business owner's photos including a cow in a field
Mike Hall, owner of a small digital photo gallery, had images of wildlife, landscapes and buildings blocked by Facebook's ad system for supposedly containing 'overtly sexual' content, including a photo of a cow. Hall appealed but received no response until the BBC contacted Facebook, after which the ads were reinstated and Facebook apologised, citing a technical error.
Company
Facebook
AI system
Facebook Ads
9 source articlesbbc.co.uk
WF-I55ELHLive 5 Oct 2026
Wisconsin’s dropout prediction algorithm labels students high risk with racial bias
Wisconsin’s Dropout Early Warning System (DEWS) uses machine learning to predict middle school students' likelihood of graduating on time, relying on factors including race and income. The Markup found the algorithm frequently mislabels students, with a higher false alarm rate for Black and Hispanic students. The Wisconsin Department of Public Instruction has not informed schools of the racial disparities and continues to use the system.
Company
Wisconsin Department of Public Instruction
AI system
Dropout Early Warning System (DEWS)
10 source articlesthemarkup.org
WF-IW6AHHHappened 1 Dec 2020Live 5 Oct 2026
Twitter bot network attacked Belgian government's Huawei 5G ban
A network of 14 Twitter accounts using GAN-generated profile images engaged in a coordinated campaign to criticize the Belgian government's plan to ban Huawei from supplying 5G equipment. The accounts posed as Belgium-based tech experts and retweeted content praising Huawei. Twitter has suspended all 14 accounts.
10 source articleszdnet.com
WF-E7VSR2Happened 1 Jan 2021Live 5 Oct 2026
LA's VI-SPDAT housing scoring system gives lower priority to Black and Latino people
The Los Angeles Homeless Services Authority uses the VI-SPDAT scoring system to prioritise unhoused people for subsidised permanent housing. An investigation by The Markup found that Black and Latino people consistently receive lower vulnerability scores than White people, leading to lower priority for housing. The agency has acknowledged the racial disparities and is working on a new tool, but continues to use the current system.
Company
Los Angeles Homeless Services Authority
AI system
VI-SPDAT
10 source articlesthemarkup.org
WF-ZEZR6OLive 5 Oct 2026
Chinese tech companies patented Uyghur detection analytics
IPVM reported that Huawei, Megvii, SenseTime, and other Chinese tech companies filed patents in China for AI systems that can detect Uyghur ethnicity. The patents included Uyghur as a detectable attribute for facial recognition and image retrieval. The companies responded by saying they would amend or withdraw the patents, claiming the references were misunderstood or regrettable, while the technology is used by police for surveillance and targeting of Uyghurs.
10 source articlesipvm.com
WF-9LKOK3Live 5 Oct 2026
Portland Metro ends Replica partnership over data privacy concerns
Portland Metro, an elected regional government in Oregon, ended its pilot project with movement data company Replica after a disagreement about data sharing. Portland Metro requested raw, disaggregated data, which Replica refused to provide, citing user privacy concerns. The partnership was terminated without payment.
Company
Portland Metro
AI system
Replica
10 source articlesgovtech.com
WF-1980JLLive 5 Oct 2026
OpenAI's CLIP vision system fooled by handwritten notes
OpenAI researchers discovered that their CLIP computer vision system can be deceived by handwritten labels placed on objects. The system's multimodal neurons respond to text as well as images, causing it to misidentify objects. The attack, called a typographic attack, is a research finding and not a deployed system. No actual harm occurred.
Company
OpenAI
AI system
CLIP
10 source articlestheverge.com
WF-10GPZYHappened 26 Oct 2019Live 5 Oct 2026
Deepfake bot submitted 1,001 comments to Idaho Medicaid waiver website
A researcher created a bot that generated and submitted 1,001 deepfake comments to the federal public comment website for the Idaho Medicaid reform waiver in October 2019. The comments were indistinguishable from human submissions, and survey respondents could not tell them apart from real comments. The researcher later withdrew the comments. The study highlights the vulnerability of federal comment processes to automated manipulation.
8 source articlestechscience.org
WF-M5XXFQLive 5 Oct 2026
Amazon apologises after Alexa plays porn to a toddler who asked for a song
A toddler named Bobby asked an Amazon Echo Dot's Alexa assistant to play the children's song 'Digger, Digger'. Instead, Alexa responded with explicit pornographic audio, prompting the family to switch the device off. Amazon apologised to the family and said the glitch had been fixed, adding that it was working on restrictions to prevent a recurrence.
Company
Amazon
AI system
Alexa
10 source articlesnypost.com
WF-TX7ZJMHappened 8 Mar 2021Live 5 Oct 2026
Verkada security breach exposes customer video and data
In March 2021, attackers compromised Verkada's platform and accessed video and image data from 97 customer organisations. The attackers used a misconfigured customer support server to gain access and viewed live video, accessed badge credentials for eight customers, and downloaded user lists. Verkada cut off access within hours and notified affected customers. The attacker, Tillie Kottmann, was later indicted by the U.S. Department of Justice.
Company
Verkada
AI system
Verkada Command platform with People Analytics
10 source articlesverkada.com

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

Dallas police used robot to kill gunman after standoff

Cruzcampo deepfake ad of Lola Flores goes viral

Kohler, BMW, MaxMara secretly collected customers' facial recognition data

AnyVision patents drone facial recognition technology

Seoul Metropolitan Government rolls out AI CCTV cameras to prevent suicides

Deepfake accounts pretending to be Amazon workers cause confusion on Twitter

DoorDash changes tipping policy after backlash over worker tips

ByteDance whistleblower alleges Chinese government uses TikTok algorithms to censor content

New York City's McKinsey-led jail violence program manipulated data, violence increased

Facebook's policy change brought partisan content back to Georgia feeds

Israeli military uses Smash Hopper turret against Palestinian protesters

Facebook's automated news ban blocks non-news pages in Australia

Facebook continued recommending political groups despite promise to stop

Serbia's Social Card law could harm marginalized groups

Facebook's ad system erroneously blocks business owner's photos including a cow in a field

Wisconsin’s dropout prediction algorithm labels students high risk with racial bias

Twitter bot network attacked Belgian government's Huawei 5G ban

LA's VI-SPDAT housing scoring system gives lower priority to Black and Latino people

Chinese tech companies patented Uyghur detection analytics

Portland Metro ends Replica partnership over data privacy concerns

OpenAI's CLIP vision system fooled by handwritten notes

Deepfake bot submitted 1,001 comments to Idaho Medicaid waiver website

Amazon apologises after Alexa plays porn to a toddler who asked for a song

Verkada security breach exposes customer video and data