AI Misconduct.com

The daily rap sheet of the machine age.

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

Tuesday, July 29, 2025

8 items

Family sues school district over response to real and AI nudes

Who did it: HUMAN ABC7 Los Angeles

A family has sued California's Las Virgenes Unified School District, alleging it failed to respond adequately to repeated reports of harassment over seven months after a classmate allegedly secretly filmed their daughter and used AI to put her face on other bodies, with some images ending up on a pornographic website. The district says it immediately reported the allegations to the sheriff's department.

Goldberg Segalla's one fake case becomes many, plaintiffs allege

Who did it: HYBRID Above the Law

Goldberg Segalla admitted that a Chicago Housing Authority motion attacking a $24.1 million lead-poisoning verdict cited Mack v. Anderson, a case supplied by ChatGPT that does not exist. Plaintiffs have now moved for sanctions. They allege further fake cases, fabricated quotations and misrepresentations in other filings, including a February motion seeking more time for legal research.

Judge finds phantom statute subsection and case in contractor's brief

Who did it: HUMAN AI Hallucination Cases database

US District Judge Charles E. Atchley Jr. dismissed a contractor's claim against First US Bank. The judge found the plaintiffs cited a nonexistent Alabama Code subsection, a case that does not exist and real cases that did not support them. Unsure whether generative AI, carelessness or intent was to blame, the judge warned counsel that repeats would bring sanctions.

OCAHO tosses complaint over AI-invented regulation and case law

Who did it: HYBRID AI Hallucination Cases database

A Justice Department administrative law judge dismissed with prejudice a pro se complaint against Droisys, Inc. The complainant's filings had cited a nonexistent regulation, an invented OCAHO decision and fabricated quotes despite an earlier warning. The complainant blamed publicly available AI tools and a lack of access to Westlaw; the judge ruled that neither was good cause.

Former staff say Builder.ai's AI work was mostly done by humans

Who did it: HUMAN Rest of World

Former Builder.ai employees told Rest of World that the collapsed startup, once valued at $1.5 billion, used AI only for relatively basic tasks. They said staff and outsourced developers in countries including India and Ukraine did most of the work. Its final CEO, Manpreet Ratia, denied the company misled people, writing: "THIS IS NOT A STORY ABOUT FAKE AI."

YouTuber Chimchakman plans legal action over deepfake gambling ads

Who did it: HUMAN Sports Kyunghyang

YouTuber Chimchakman, who has 2.89 million subscribers, warned on the 29th that some companies are using AI to synthesise the creator's face and voice in illegal gambling game ads. The notice says Chimchakman never made such ads. It plans strict legal action over publicity and portrait rights, copyright and unfair-competition violations.

Deepfake clips falsely show Romanian PM announcing war in September

Who did it: HUMAN Digi24

Romania's Interior Ministry warned of a disinformation campaign spreading AI-made videos in which Prime Minister Ilie Bolojan appears to announce that Romania will enter a war from September 2025. The ministry called the clips deliberate manipulation of public opinion and said Bolojan made no such statements. It urged people not to share unverified content.

Deepfake Jorge Ramos falsely reports Melania and Barron deportation order

Who did it: HUMAN El Diario NY

Viral videos claiming the US Supreme Court ordered the deportation of Melania and Barron Trump were created with AI, according to independent fact-checkers. The clips use journalist Jorge Ramos's image without consent. Hiya Deepfake Voice Detector found the voice was likely synthetic. No such ruling exists, and Melania Trump became a US citizen in 2006.