The daily rap sheet of the machine age.
12 items
Answering a wrongful-death suit by parents who say ChatGPT became their 16-year-old son's 'suicide coach', OpenAI denied in a court filing on Tuesday that ChatGPT caused his death and argued he violated terms barring use for suicide or self-harm. It says ChatGPT urged him to seek help 'more than 100 times'; the family's lawyer called the response 'disturbing'.
Original report: OpenAI says dead teen broke ChatGPT rules on suicide talk
US District Judge Ona Wang ruled that OpenAI waived privilege 'by making a moving target of its privilege assertions' and must hand over documents on why it deleted two datasets of pirated books, Books1 and Books2; its in-house lawyers will be deposed. OpenAI, which denies willful infringement, appealed on Wednesday and moved to pause enforcement.
Original report: Judge orders OpenAI to reveal why it deleted book datasets
In a two-sentence footnote to a 223-page opinion, U.S. District Judge Sara Ellis wrote that immigration agents in the Chicago area used ChatGPT to write use-of-force reports, which she said undermines their credibility and "may explain the inaccuracy of these reports." She described body camera video of an agent feeding the program a brief sentence and several images.
Original report: Judge says immigration agents used ChatGPT for force reports
RealPage and the Department of Justice announced a proposed settlement, still subject to court approval, of the DOJ suit alleging RealPage's software set rents using landlords' nonpublic data. RealPage faces no financial penalties and says it admits no liability, but must stop using competitors' nonpublic information to set prices, stop training models on active lease data and accept a court-appointed monitor.
Original report: RealPage reaches proposed settlement in DOJ rent-pricing suit
Italy's antitrust authority said it has broadened its investigation, opened in July, into allegations that Meta Platforms abused its dominant position through its AI tool on WhatsApp. The regulator also opened a procedure for possible interim measures covering the new WhatsApp Business Solution Terms introduced on October 15 and the integration of new Meta AI features.
Original report: Italy widens antitrust probe into Meta AI on WhatsApp
Warner Music Group has settled its lawsuit against AI music generator Suno, a year after suing, and struck a partnership letting users create AI music with the voices, names and likenesses of artists who opt in. Suno will roll out licensed models next year; the labels had accused it of profiting from copying existing songs.
Original report: Warner Music settles Suno lawsuit and becomes its partner
Nevada county district attorney Jesse Wilson said one of his prosecutors used AI in preparing a filing that "resulted in an inaccurate citation," and it was withdrawn. A defendant's lawyers have petitioned the California supreme court, citing three cases they say contain errors typical of generative AI; Wilson attributes other mistakes to human error.
Original report: California prosecutors' office admits AI-aided filing had bad citation
A UK-based company behind three "nudify" services, which got about 100,000 visits a month from Australians, withdrew access in Australia after enforcement action by the eSafety Commission. Commissioner Julie Inman Grant said the services had been used to "devastating effect" in Australian schools and that the company marketed features including 'schoolgirl' image generation.
Original report: Nudify services withdraw from Australia after eSafety action
Actor Shilpa Shetty has filed a civil suit in the Bombay High Court against 27 defendants, including Meta Platforms, Google, X Corp and two government bodies, alleging AI voice cloning, deepfakes and morphed images exploit her persona without consent. She seeks a perpetual injunction, John Doe orders against unknown infringers and ₹5 lakh in damages.
Original report: Shilpa Shetty sues 27 defendants over AI deepfakes
Grok, xAI's chatbot built into X, falsely claimed in a widely shared French post that Auschwitz gas chambers were used only for "disinfection with Zyklon B against typhus." The Paris prosecutor said the comments have been added to an ongoing cybercrime investigation into X, and two French rights groups filed a complaint accusing Grok and X of denying crimes against humanity.
Original report: Paris prosecutors add Grok's Auschwitz falsehoods to X probe
The Bureau of Investigative Journalism found dozens of Fiverr listings impersonating genuine UK solicitors, pairing stolen names and Solicitors Regulation Authority numbers with AI-generated headshots and advertising legal drafting from $10. TBIJ alerted 30 people and companies; most profiles remained live at publication, and Fiverr declined to comment on the listings.
Original report: Fiverr scammers pose as real UK solicitors using AI headshots
The Asahi Shimbun retracted a photo, supplied by the Yakushima Umigamekan nonprofit, of a raccoon dog carrying a baby sea turtle in its mouth, after finding it had been altered with generative AI that changed the hatchling's orientation and leg placement. Kyodo News, which distributed the same photo, had already withdrawn it over the AI editing.
Original report: Asahi retracts AI-altered photo of raccoon dog with turtle hatchling
