AI Misconduct.com

The daily rap sheet of the machine age.

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

Wednesday, July 30, 2025

6 items

Italy's antitrust watchdog investigates Meta over WhatsApp AI

Who did it: HUMAN Reuters

Italy's antitrust authority, AGCM, said on July 30 it had opened an investigation into whether Meta abused its dominant position by building its Meta AI assistant into WhatsApp without user consent, potentially harming competitors, and had carried out investigations at the offices of Meta's Italian subsidiary. Meta says it is cooperating and that free access gives "millions of Italians" the choice to use AI.

Prescott eNews runs racist AI cartoons of Rep. Quang Nguyen

Who did it: HUMAN Arizona Mirror

Prescott eNews, a local site owned by former Arizona lawmaker David Stringer, published AI-generated "political cartoons" depicting Republican Rep. Quang Nguyen, whose family fled Vietnam, as a police suspect in stealing and eating neighborhood pets and questioning his loyalty and citizenship, drawing criticism from Stop AAPI Hate and Facebook commenters. Nguyen called the attacks "vile"; Stringer did not respond to requests for comment.

Court sanctions Apple challengers over ChatGPT petition and fake citations

Who did it: HYBRID AI Hallucination Cases database

Judge Edward M. Chen ruled on July 30 that plaintiffs suing Apple and their lawyer, Keith Mathews, warrant Rule 11 sanctions over a res judicata-barred suit, a ChatGPT-written petition demanding Tim Cook's resignation filed three times, and an anti-SLAPP motion citing a non-existent case. Ironically, the court noted, their show-cause response contained a hallucinated quote; Apple is entitled to reasonable fees.

Mexico City police warn of AI scams targeting car-theft victims

Who did it: HUMAN El Universal

Mexico City's cyber police unit warned that fraudsters are using AI-generated messages, emails and calls to pose as authorities, helpful citizens or tow-truck staff, falsely telling car-theft victims their vehicle has been found and demanding rewards or transport fees. The scammers draw on details victims post on social media, such as plates, model and where the theft happened.

Judge warns pro se Amazon litigant over nonexistent citations

Who did it: HYBRID AI Hallucination Cases database

U.S. District Judge John H. Chun on July 30 warned a pro se plaintiff suing Amazon that further filings with fake citations could bring sanctions, after the plaintiff's briefs cited a nonexistent case, whose Westlaw number led to a Florida court filing, and a made-up quote from Hall v. Hall. The judge said Amazon's suggestion that AI drafted the briefs "may have merit".

ByteDance's Trae IDE keeps phoning home after telemetry opt-out

Who did it: HUMAN The Register

An analysis of ByteDance's AI coding editor Trae found it made around 500 calls within about seven minutes, transferring up to 26 MB of data including machine identifiers, user IDs and project information, after telemetry was supposedly disabled. ByteDance has since reworded the setting to say it only covers telemetry via the VS Code framework, not other Trae tools.