AI Misconduct.com

The daily rap sheet of the machine age.

Updated Nov 2, 2026
4,173 items in the archive
RSS feed

Tuesday, October 7, 2025

8 items

OpenAI bans ChatGPT users drafting surveillance pitches linked to China

Who did it: HUMAN CNN

A ChatGPT user likely linked to a Chinese government entity sought help writing a proposal for a tool analysing travel movements and police records of Uyghurs and other "high-risk" people, and another Chinese-speaking user sought promotional materials for a social media scanning tool, OpenAI reported; it banned both. China's embassy said it opposes "groundless attacks and slanders".

Philippine privacy watchdog orders Orb maker to stop collecting biometrics

Who did it: HUMAN Newsbytes.PH

The Philippines' National Privacy Commission ordered Tools for Humanity, maker of the World App and Orb verification system, to stop collecting and processing personal and biometric data for allegedly violating the Data Privacy Act. The commission said paying people to undergo iris scans was "undue influence" invalidating consent, and the collection "excessive and disproportionate".

Lucknow police arrest exam gang accused of AI face-blending

Who did it: HUMAN The Times of India

Lucknow police arrested 10 members of a solvers' gang, including alleged mastermind Anand Kumar, an assistant manager at Prathama UP Gramin Bank. Police say the gang used AI-based apps to blend photos of real candidates and proxies, backed by forged Aadhaar and PAN cards. It unravelled when invigilators grew suspicious of a proxy at the IBPS Clerk Examination 2025.

Woman tells court deepfake Abela videos cost her €53,250

Who did it: HUMAN timesofmalta.com

A woman told a Maltese court she paid a total of €53,250, "practically all my life savings", into a trading scheme after seeing deepfake Facebook videos of Prime Minister Robert Abela purportedly promoting it. She testified against Ukrainian national Kateryna Izotkina, arrested after a police controlled delivery and accused of money laundering, fraud and organised crime, which she denies.

Alabama homeowners sue State Farm alleging 'cheat and defeat AI'

Who did it: HYBRID Insurance Business

Two elderly, disabled Black homeowners have sued State Farm in federal court in Alabama, alleging its claims processing relied on "cheat and defeat AI algorithms as discriminatory tools" that subjected their claims to greater scrutiny than white policyholders'. They seek $372,437.36 plus interest over lightning and water damage; no findings of fact have been made.

Austria's top court rejects appeal with apparently AI-generated false citations

Who did it: HYBRID AI Hallucination Cases database

Austria's Supreme Court rejected a convicted drug trafficker's nullity appeal on 7 October 2025, noting part of the filing was riddled with false citations, including to Supreme Court decisions mostly nonexistent or not on the stated topic, apparently produced by "so-called artificial intelligence" without professional review. The court declined to address its substance.

Fake AI news clip borrows Hungarian lottery's name to push app

Who did it: HUMAN Index

Hungary's national lottery company Szerencsejáték Zrt. warned that a misleading AI-made video disguised as a TV news report is circulating on social media, suggesting it recommends downloading an app. The company says the advertiser, Szerencse Klub, falsely implies a link to the lottery operator, and that it will use every legal option against such abuses.

Apple seeks dismissal of class action over delayed Siri features

Who did it: HUMAN Gadget Hacks

Apple has filed a motion to dismiss a class action over delayed Apple Intelligence features. It argues that two missing Siri features do not justify sweeping legal claims, and that it said features would be 'delivered over time and continue to evolve.' Separately, a Los Angeles class action accuses Apple of misrepresenting iPhone 16 AI features, and shareholders are suing over the delays.