Common Sense Media says hour-long conversations about suicidal ideation, self-harm or disordered eating on newly created, parent-linked ChatGPT for Teens accounts produced "zero" parental alerts, and the bot failed to reliably point teens in crisis to a hotline or professional. It called the product an "unacceptable risk"; OpenAI says much of the testing may have predated full activation of parental controls.
Students and parents protested outside Colegio Modelo Lomas in Lomas de Zamora after pupils allegedly used AI to fake nude images, and some videos, of female classmates, reportedly spreading them on porn sites; estimates of victims range from 30 to about 50. A juvenile prosecutor is investigating, and two adolescent pupils have been notified of the case.
Kokomo police officer Jeffrey Bornes Jr., 40, has been charged with Level 6 felony official misconduct and two misdemeanor fraud counts; prosecutors allege he used his department Flock license-plate-reader account 18 times to track his then-spouse's vehicle, logging each search as a "welfare check". The department found the searches in its own audit of over 9,000 entries.
A Perth woman says men at a Scarborough venue secretly photographed her and a friend, used AI to make sexualised images of them and showed them around, laughing. The venue banned one man, and a detective told the women "there is no law" on AI-created media, though WA law covers creating intimate images without consent.
Coroner Janet Anderson found the actions of "Donna Chase", whom police believe was an AI-generated persona, were "the primary relevant factor" in a 41-year-old Auckland man's 2024 suicide, hours after the persona solicited sexual images over Signal and demanded money and Steam gift card codes. Police believe the account he paid is tied to an overseas criminal network.
The Malaysian Communications and Multimedia Commission (MCMC) confirmed it received a complaint that AI was misused to create and distribute sexual images and videos allegedly involving students and teachers at a school in Kedah. It will investigate with police and the communications, home and education ministries, noting penalties of up to RM500,000 or two years' jail, and up to five years for offences against children.
An unidentified developer used the GTA V source code leaked in late 2023 and AI tools to port the PC game to WebAssembly so it ran in Chrome, multiplayer included, according to Gizmodo. The site vanished after a day, leaving only "Thanks for playing!"; several outlets attributed this to a Take-Two cease-and-desist letter, and the company has not responded.
A preliminary IMDEA Networks Institute study says all nine chatbot services it examined, including ChatGPT, Claude, Gemini, Grok and Meta AI, pass user data to third parties. Five shared conversation permalinks with trackers, often without access restrictions, and three passed auto-generated chat summaries, some covering health or salary questions, to ad partners, with recipients including Meta, TikTok and Google, the researchers say.
Florida's 4th District Court of Appeal found Winsor Daniel Jr.'s brief for Kid International cited the non-existent Ogilvie v. Sebring Airport Authority seven times, fictional quotation included, and misquoted real cases. Daniel admitted the authorities were "not adequately verified", pointing to client-supplied research and his firm's generative AI use; the court affirmed summary judgment against his clients and took up the defendants' sanctions motion.
Actor Vijay Deverakonda has asked the Delhi High Court to protect his personality rights against digital manipulation of his face and voice, fake endorsements, people allegedly posing as his agents and merchandise using his likeness. The court issued summons and notice on his interim injunction plea, and Justice Vikas Mahajan said an order will be passed; his lawyers said offending URLs were removed after filing.
