AI Misconduct.com

The daily rap sheet of the machine age.

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

Thursday, October 2, 2025

5 items

Citizen Lab exposes AI-fuelled X network urging Iranian revolt

Who did it: HUMAN The Citizen Lab

Citizen Lab says a network of more than 50 inauthentic X profiles it calls PRISONBREAK used AI-generated content, including a fake video of the June 23 Israeli strike on Evin Prison that news outlets republished, to urge Iranians to revolt. It assesses an Israeli government agency or close subcontractor as most likely behind it, but cannot conclusively attribute it.

Videos attacking Péter Magyar, often AI-made, ran over 650 million times

Who did it: HUMAN hvg.hu

Videos attacking Péter Magyar and the Tisza Party, often made with AI, ran more than 650 million times on Facebook and YouTube in three months, Telex reports. They come from the Nemzeti Ellenállás Mozgalom, founded by Fidesz-aligned pundit Bence Apáti, which paid 643 million forints to run them, according to Google and Meta ad databases.

Singapore lawyer must pay opponent $800 over AI-invented case

Who did it: HYBRID The Straits Times

A Singapore lawyer, Lalwani Anil Mangan of DL Law Corporation, was ordered to pay $800 to his client's opponent after submissions cited a non-existent case that he said a junior lawyer had likely run through an AI app. Assistant Registrar Tan Yu Qing called his conduct improper, unreasonable and negligent, and said he was less than candid.

Voice actors call Seoul Metro's AI voice-cloning plan illegal

Who did it: HUMAN Yonhap News Agency

Seoul Metro's plan to replicate with AI, without her consent, the voice of Kang Hee-sun, who has voiced subway station announcements for 29 years, is illegal under copyright law and ethically blameworthy, the Korea Voice Actors Association said on Oct. 2 in a joint statement, demanding a formal apology and explicit consent in future.

Lawyer sanctioned $1,000 in Louisiana federal court for AI-fabricated citation

Who did it: HYBRID AI Hallucination Cases database

US District Judge Susie Morgan sanctioned lawyer Pius A. Obioha $1,000, payable personally, after his opposition brief cited a fabricated, AI-generated case and misstated another case's holding. Obioha admitted he signed it without checking the citations; he must take an hour of generative-AI CLE by January 31, 2026, and was referred for discipline.