Where automated decisions went wrong

Jordan's National Aid Fund Algorithm Denies Cash Assistance to Vulnerable Households

Human Rights Watch reported that Jordan’s Takaful cash transfer programme, run by the National Aid Fund and financed by the World Bank, uses an algorithm to rank households by estimated economic vulnerability and select recipients. The report found that the algorithm relies on inaccurate self-reported income and questionable indicators such as car ownership and utility use, causing eligible poor families to be denied benefits or receive reduced payments. The World Bank and NAF defended the programme and its targeting approach.

Company involved

National Aid Fund (NAF)

AI system involved

Takaful (Unified Cash Transfer Program)

3 source articles · read the reporting →

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 involved

Idaho Department of Health and Welfare

10 source articles · read the reporting →

INSS AI Denies Rural Worker's Pension After Misidentifying Her as a Man

A rural worker in Brazil, Josélia de Brito, had her pension application through the Meu INSS app automatically denied after the AI system misidentified her as a man. The INSS deployed the system to speed up benefit decisions, but experts say it struggles with complex rural cases. The case is cited as evidence that automation may exclude vulnerable people with limited digital access.

Company involved

Instituto Nacional do Seguro Social (INSS)

AI system involved

Meu INSS

4 source articles · read the reporting →

SenseNets silent after data leak exposes millions of people's records

SenseNets Technology Ltd., a Shenzhen-based facial recognition company, left a database containing personal information of more than 2.5 million people publicly accessible without password protection for months. Dutch security researcher Victor Gevers and the GDI Foundation discovered the exposure in July and warned the company, which did not respond. The database was secured in February after the leak was reported, and the company is reported to be conducting an internal investigation. SenseNets has declined to comment publicly.

Company involved

SenseNets Technology Ltd.

10 source articles · read the reporting →

France: CNAF's discriminatory risk-scoring algorithm must be stopped

Amnesty International and coalition partners filed a complaint with the Council of State against CNAF's risk-scoring algorithm used to detect benefit overpayments. The algorithm assigns risk scores based on criteria that discriminate against vulnerable groups, including those with disabilities, single parents, and low-income households. The complaint alleges the system violates human rights to equality and privacy. The EU AI Act's social scoring ban may apply, but its definition remains unclear.

Company involved

CNAF

2 source articles · read the reporting →

AMS algorithm lacks transparency and may discriminate against job seekers

The Austrian Public Employment Service (AMS) uses an algorithm to classify job seekers into categories A, B, and C, determining their access to benefits and training. Scientists from TU Wien, WU Wien, and University of Vienna have criticised the algorithm for lacking transparency, as only two of 96 model variants have been published. They allege that the system may discriminate against women and people with migration background, and that job seekers are not informed about how the algorithm works or given a chance to appeal.

Company involved

AMS (Arbeitsmarktservice Österreich)

AI system involved

AMS-Algorithmus

10 source articles · read the reporting →

Cense exposed 2.5 million records of auto accident victims online

On July 7, 2020, a security researcher discovered 2.5 million records containing personal and medical data of auto accident victims exposed online. The records, belonging to New York-based AI company Cense, included names, insurance policy numbers, claim numbers, and medical diagnosis notes. The data was labeled as staging data, possibly intended for temporary storage before being loaded into an AI system. After the researcher sent a responsible disclosure notice, Cense restricted public access to the database.

Company involved

Cense

AI system involved

Cense

5 source articles · read the reporting →

CanLII sues Caseway AI for scraping legal database

The Canadian Legal Information Institute (CanLII) has filed a lawsuit in British Columbia Supreme Court against Caseway AI, alleging that the company's AI chatbot scraped approximately 3.5 million records from CanLII's database in bulk, violating its terms of service and copyright. CanLII claims it adds value to public court records through hyperlinks and corrections, which it says constitute protected copyrighted work. Caseway AI argues the information is public and accessible elsewhere, and that it did not use CanLII's enhancements. The lawsuit was settled in March 2026, with terms undisclosed.

Company involved

Canadian Legal Information Institute (CanLII)

AI system involved

Caseway

5 source articles · read the reporting →

California bar exam facial recognition system fails to verify Arab-American student's identity

Ahmed Alamri, an Arab-American law student, was unable to register for the practice California bar exam because ExamSoft's facial recognition system repeatedly failed to recognize his face, citing poor lighting. Alamri attempted to verify his identity over 75 times in different rooms and lighting conditions without success. He and other students filed an emergency petition with the California Supreme Court, alleging that the system discriminates against people of color. The incident highlights concerns about bias in facial recognition technology used for exam proctoring.

Company involved

California State Bar

AI system involved

ExamSoft

10 source articles · read the reporting →

DeepScore markets facial and voice analysis app for trustworthiness scoring despite experts' doubts

DeepScore, a Tokyo-based company, is marketing an app that uses facial and voice recognition to score people's trustworthiness for lenders and insurers in Japan, Indonesia, Vietnam and the Philippines. The company says the app can detect deception with 70 per cent accuracy, but researchers and privacy advocates say there is no reliable scientific basis for such judgments and warn of discrimination and privacy harms. The chief executive said the system is only one part of lenders' and insurers' decision-making and that people can choose not to use it. Critics respond that an unequal balance of power makes consent difficult.

Company involved

DeepScore

AI system involved

DeepScore

6 source articles · read the reporting →

CBSE introduces facial recognition system for students to access digital documents

The Central Board of Secondary Education (CBSE) has introduced a facial recognition system for Class 10 and 12 students to access their digital academic documents. A live image of the student is compared with the photograph on their CBSE admit card and, if the match succeeds, the certificate is emailed to them. The facility is available on Digi Locker for 2020 records and is expected to help foreign students and those unable to open a Digi Locker account.

Company involved

Central Board of Secondary Education (CBSE)

AI system involved

Facial Recognition System

10 source articles · read the reporting →

Lemonade backtracks on tweet claiming AI scans customer faces for fraud

Lemonade, an insurance firm, posted a tweet claiming it uses AI to detect non-verbal cues in customer videos to deny claims. After backlash comparing the practice to phrenology, the company deleted the tweet and denied using AI to deny claims based on facial characteristics. Lemonade stated it uses facial recognition only to flag claims submitted under different identities, which are then reviewed by human investigators.

Company involved

Lemonade

10 source articles · read the reporting →

Joann LeDoux v. Outliers, Inc. (2) (W.D. Washington): AI-hallucinated content in court filing, Expert Brief excluded/struck

The AI-generated hallucinated citations in an expert report led to the exclusion of the expert and dismissal of the plaintiff's case.

1 source article · read the reporting →

Denmark's Udbetaling Danmark welfare algorithm illegally surveilled millions

Udbetaling Danmark (UDK), Denmark's centralized welfare payment agency, used automated algorithms to check eligibility and detect fraud. The system collected personal data on millions of beneficiaries and their relatives, including non-beneficiaries, in what a think-tank called 'systematic surveillance'. A complaint to the Danish Data Protection Authority led to a ruling that the data collection violated GDPR, and UDK eventually deleted the illegally obtained data. The system also caused erroneous payments, including underpaying 300,000 pensioners by 94 million euros annually.

Company involved

Udbetaling Danmark

10 source articles · read the reporting →

Intellexa Predator spyware used to surveil human rights lawyer in Pakistan

In summer 2025, a human rights lawyer from Pakistan's Balochistan province was targeted via WhatsApp with Intellexa's Predator spyware, according to Amnesty International's Security Lab. The attack is part of a broader pattern of unlawful surveillance of activists, journalists and human rights defenders. The Intellexa Leaks investigation exposed internal operations of the spyware company, raising concerns about human rights due diligence.

AI system involved

Predator spyware

10 source articles · read the reporting →

LAPD told officers to collect social media data on every civilian stopped

The Los Angeles Police Department directed officers to record the social media accounts of every civilian they interviewed, including those not arrested or accused of a crime, according to records obtained by the Brennan Center for Justice. The field interview cards, which also ask for social security numbers, have raised civil liberties concerns. Prosecutors have accused three officers of using the cards to falsely label civilians as gang members. The LAPD said the policy was being updated but did not respond to questions about its use of social media monitoring tools including Media Sonar and Geofeedia.

Company involved

Los Angeles Police Department (LAPD)

AI system involved

Field interview cards, Media Sonar, Geofeedia

10 source articles · read the reporting →

UnitedHealth used ALERT algorithm to limit mental health coverage, regulators found

ProPublica reports that UnitedHealth Group's Optum subsidiary used the ALERT algorithm to flag mental health patients and therapists as outliers, leading to therapy coverage denials. Regulators in California, Massachusetts and New York alleged this breached the federal Mental Health Parity and Addiction Equity Act, and UnitedHealth agreed to restrict the system in those jurisdictions. The company denies wrongdoing and says its programmes are compliant. Affected patients, including Medicaid enrollees, were left to pay out-of-pocket or go without care.

Company involved

UnitedHealth Group

AI system involved

ALERT

5 source articles · read the reporting →

IRCC uses AI triage for Temporary Resident Visa applications

Immigration, Refugees and Citizenship Canada (IRCC) uses an AI system called Advanced Analytics to triage Temporary Resident Visa applications from India and China. The system categorizes applications into tiers, with Tier 1 approved automatically and others sent to human officers. Critics allege the system lacks transparency and may introduce bias, leading to visa refusals without clear rationale. The author, a Canadian immigration lawyer, is filing Federal Court cases on behalf of clients affected by refusals.

Company involved

Immigration, Refugees and Citizenship Canada (IRCC)

AI system involved

Advanced Analytics Triage of Overseas Temporary Resident Visa Applications

10 source articles · read the reporting →

Five Illinois residents sue facial recognition company Pimeyes under BIPA

Five Illinois residents are suing Pimeyes, a facial recognition search engine, alleging that the company violated the Illinois Biometric Information Privacy Act by collecting and using their facial data without consent. The lawsuit claims that Pimeyes made the residents' photos and information available in search results, enabling unauthorized access. The company has not yet been served with the lawsuit, and its CEO stated he had no prior knowledge of the case.

Company involved

Pimeyes

AI system involved

Pimeyes

7 source articles · read the reporting →

California EDD's automated fraud detection wrongly suspended 600,000 legitimate unemployment claims

In January 2021, the California Employment Development Department used Thompson Reuters automated batch review software to flag 1.1 million unemployment claims as potentially fraudulent. EDD stopped payments on those claims without prior notice. Later, over 600,000 were confirmed as legitimate after claimants used ID.me to verify their identity. The incident highlights the trade-off between fraud prevention and timely benefit access.

Company involved

California Employment Development Department (EDD)

AI system involved

Thompson Reuters Automated Batch Review

7 source articles · read the reporting →

OnlyFake site uses neural networks to generate fake IDs, bypasses OKX verification

An underground website called OnlyFake uses neural networks to generate realistic photos of fake IDs for $15. The journalist tested the service and obtained a convincing California driver's license. They then used another fake ID to successfully bypass the identity verification process on OKX, a cryptocurrency exchange. The article alleges that this technology could streamline bank fraud and money laundering, but reports no actual financial loss.

Company involved

OKX

AI system involved

OnlyFake

10 source articles · read the reporting →

Swedish welfare agency's AI system flags marginalized groups for fraud investigations

Försäkringskassan, Sweden's Social Insurance Agency, uses an AI risk-scoring system to flag welfare applicants for fraud investigations. The system disproportionately targets women, individuals with foreign backgrounds, low-income earners, and those without university degrees, according to an investigation by Lighthouse Reports and Svenska Dagbladet. Amnesty International has called for the system to be discontinued, citing violations of the right to equality and non-discrimination.

Company involved

Försäkringskassan (Swedish Social Insurance Agency)

6 source articles · read the reporting →

iBorderCtrl lie detector falsely flagged honest reporter as liar

A journalist testing Europe's iBorderCtrl virtual policeman at the Serbian-Hungarian border gave honest answers but was deemed a liar by the system, scoring 48 out of 100 with four false answers flagged. The Hungarian policeman said the result suggested further checks, though none were carried out. The reporter only learned of the result after filing a data access request under European privacy laws. Experts and transparency activists have criticised the technology as pseudoscientific and potentially discriminatory.

Company involved

iBorderCtrl consortium

AI system involved

Silent Talker / iBorderCtrl virtual policeman

10 source articles · read the reporting →

Woman denied state welfare card groceries after face scan fails

A grandmother rode her motorcycle several kilometres to a shop in Sapphaya district, Chai Nat province, to use her state welfare card to buy groceries for her family. After selecting items, she failed repeated facial recognition scans because her ID card photo was years old and no longer matched her appearance. She left the shop empty-handed and in tears.

Company involved

Government of Thailand (state welfare card system)

2 source articles · read the reporting →