The daily rap sheet of the machine age.
10 items
A Tennessee woman jailed for months in a Fargo bank fraud case, later dismissed, has filed a $10 million federal civil rights suit against Fargo police and Detective Lucas Heck. The complaint alleges police treated an AI facial recognition hit as positive identification and that Heck ignored exculpatory evidence; she was released on Christmas Eve with $0.04.
Original report: Woman jailed over AI face match sues Fargo for $10M
Universal Music Group has sued DistroKid in federal court, alleging the distributor floods Spotify and other services with "AI-generated slop" passed off as human-made, including one "artist", Lofi Chill, that released 4,562 tracks in twelve months. The suit also alleges copyright infringement and deceptive trade practices; DistroKid says it strongly disagrees and will defend itself vigorously.
Original report: UMG sues DistroKid over AI 'slop' passed off as human
The Advertising Standards Authority has banned five ads on Meta platforms for AI companion and video-generation apps, including Nexaipic, Rusto AI and PictoPop, for promoting sexually explicit content and objectifying women; one ad, for an Animcha Ltd companion app, portrayed someone under 18 sexually. Meta had already removed the ads, and none of the app makers answered the watchdog's enquiries.
Original report: UK watchdog bans AI app ads that sexualised women
A Maranhão court has ordered Google Brasil to pay R$5 million in collective moral damages over false Covid-19 health videos on YouTube, finding a "qualified omission" in its duty of safety under consumer law. The plaintiff, Ibedec-MA, argued engagement-driven algorithms spread the content; the judge rejected demands for pre-filtering or requiring creators to hold medical degrees.
Original report: Brazilian court hits Google with R$5M over YouTube health misinformation
Tasmania's Supreme Court has set aside a parole condition barring Susan Neill-Fraser from telling media she is innocent, after the court heard the board relied on a document by its secretary, Ben Hancock, containing erroneous AI information and cases that did not exist. The board did not contest; Attorney-General Guy Barnett says he will seek an explanation.
Original report: Parole gag lifted after board's rationale cited nonexistent cases
A Data Privacy Brasil and Aláfia Lab study of 413 AI-made electoral posts from 1 January to 15 August found 260 (63%) lacked the AI disclosure the TSE requires, and 81% were classed as deepfakes. Political actors posted 35%, led by Flávio Bolsonaro and Eduardo Bolsonaro with 13 each; of 53 party posts, 28 came from PL.
Original report: Brazilian AI election posts mostly unlabelled, study finds
Parkview High School in Gwinnett County, Georgia, went into lockdown on Monday after its fourth AI-generated bomb threat of the school year, which school police say fits a pattern of threats using a masked phone number. District police are working with the GBI and FBI; no arrests or suspects have been announced, and the school returned to normal operations.
Original report: Georgia high school hit with fourth AI-generated bomb threat
A Western District of Pennsylvania judge found that attorney Sharon L. Wigle violated Rule 11(b) and her duty of competence after an opposition brief bearing traces of generative AI drafting software offered six deficient citations and thirteen bare "Source" hyperlinks. The court found non-monetary sanctions appropriate.
Original report: Lawyer's AI-assisted brief earns sanctions, not a fine
A man living in Jiashan, Zhejiang, has sued the company behind the Doubao AI app, seeking an apology and compensation after it recommended 19 April as an auspicious burial date for his mother, a date he later discovered was not auspicious. Doubao, he says, also wrote his complaint and explained how to sue; Jiashan County People's Court has heard the case.
Original report: Man sues Doubao's maker using a complaint Doubao wrote
A federal judge in the Eastern District of Missouri dismissed a pro se plaintiff's case for Rule 11 violations and disobeying court orders, after a brief cited three nonexistent cases, one bearing the reporter citation of an unrelated insurance class action. The court said the citations suggest AI use, though they were hallucinations either way.
Original report: Pro se suit dismissed over three cases that never existed
