AI Misconduct.com

The daily rap sheet of the machine age.

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

Wednesday, October 29, 2025

9 items

Green Building Initiative told to explain 'totally fake' case citation

Who did it: HYBRID Reason Magazine

In Green Building Initiative v. Peacock, Judge Michael Simon (D. Ore.) found GBI cited Stell v. Cardenas, a "totally fake" case, and a Page v. Parsons citation that is only "almost real," saying they appear to be hallmarks of generative AI. GBI has two weeks to show cause why it and its counsel should not be sanctioned, and to propose its own penalty.

Hungary's Magyar to file complaint over alleged Orban-aide deepfake

Who did it: HUMAN Reuters

Hungarian opposition leader Peter Magyar said on October 29 he would file a criminal complaint against Balazs Orban, chief political aide to Prime Minister Viktor Orban (no relation), over an unlabelled 38-second Facebook video that Magyar says is an AI deepfake of him saying he would cut pensions. Balazs Orban accused Magyar of wanting to "silence those who uncover their real intentions."

Honduran lawmaker calls attorney general's audios of him AI fakes

Who did it: HUMAN Diario La Tribuna

Tomás Zambrano, the National Party's parliamentary bloc leader in Honduras, says audio recordings revealed by Attorney General Johel Zelaya are "false, altered, manipulated" with artificial intelligence and not his voice. He accused Marlon Ochoa of committing a crime by manipulating audios he presented to the Public Ministry, and said the recordings will be taken to international experts.

Lawyer says French teen in terror plot case used ChatGPT

Who did it: HUMAN AFP

A 17-year-old high school student jailed in early September on terrorist conspiracy charges used ChatGPT to help prepare plans to attack a school and a ministry or public building in Paris, his lawyer Dylan Slama said on 29 October. Slama said AI "had an influence" on his client, who he said has admitted the facts, and that adults were probably involved.

Meta attributes porn torrents to 'personal use', not AI training

Who did it: HUMAN torrentfreak.com

Seeking dismissal of a suit by adult film producers Strike 3 Holdings and Counterlife Media, which allege Meta downloaded at least 2,396 of their films since 2018 to aid AI training, Meta argues that roughly 22 downloads a year across dozens of its IP addresses are "plainly indicative of private personal use", perhaps by employees or visitors. The producers could seek up to $359 million.

Venice man arrested over 900-plus AI-generated child abuse images

Who did it: HUMAN Il Fatto Quotidiano

Italian state police have arrested a 52-year-old Venice man accused of producing more than 900 child sexual abuse images with AI tools, images investigators describe as hard to distinguish from real photos. The probe began with a tip from an international child-protection organisation; investigators say a search found an advanced system for generating such images.

Peachtree City man faces 49 counts over AI-altered images of minor

Who did it: HUMAN The Citizen

Peachtree City police say Justin Maloy, 42, used AI apps to alter photos of a minor into pornographic images; he faces 49 felony counts of sexual exploitation of a child, one per image. A chief magistrate judge found probable cause for the warrants by October 10; Maloy was transferred to Fayette County Jail on October 24 and bonded out on October 26, 2025.

ChatGPT summaries garble seven judgments in Danish Supreme Court pleading

Who did it: HYBRID AI Hallucination Cases database

A defence lawyer's pleading to Denmark's Supreme Court misdescribed seven real judgments, attributing nonexistent deportation and human-rights analyses, turning a Slovak tourist into a long-term resident with family in Denmark, and citing Üner v. the Netherlands for the opposite of its holding. The lawyer said an employee used ChatGPT for the summaries and accepted responsibility; no sanction was reported.

Child welfare agency hired ex-teacher accused of AI abuse videos

Who did it: HUMAN actionnews5.com

Wilson Jones, a former Mississippi teacher facing federal charges over AI-made sexual videos depicting students, was later hired by the state Department of Child Protection Services, court documents show, after asking the agency not to contact his former employer. Ex-superintendent Edward Lee Childress, charged with failing to report, is to be tried alongside him on November 17.