The daily rap sheet of the machine age.
9 items
An AI-generated video showing three tigers on the campus of a madrasa in Barasat went viral, panicking guardians and causing a sharp drop in attendance. The headmaster said the geography teacher made it; the madrasa issued a show-cause notice, and the teacher apologised, saying it was meant to teach students that not everything on social media is real.
Original report: AI tigers on madrasa campus earn teacher a show-cause notice
Spain's Guardia Civil is investigating a minor suspected of using AI software to turn underage schoolmates' clothed social-media photos into realistic nude images and share them with other minors. Operation 'Alhsex', opened in May after several girls reported the images, has identified five victims; police say the number may rise, and the suspect faces corruption-of-minors and revelation-of-secrets proceedings.
Original report: Guardia Civil investigates minor over AI nudes of schoolmates
China's Cyberspace Administration says it interviewed Nvidia on July 31, 2025, demanding the company explain the security risks of suspected backdoor vulnerabilities in H20 AI chips sold to China and submit supporting materials, citing claims of 'tracking and positioning' and 'remote shutdown' technology. Nvidia told The Register its chips have no 'backdoors' giving anyone remote access or control.
Original report: China summons Nvidia over alleged backdoors in H20 chips
Moldova's Central Electoral Commission says an AI-made video in which its president, Angelica Caraman, appears to say French interference in Moldova's elections is not foreign interference is a deepfake meant to manipulate opinion and undermine trust in democratic institutions. The clip spread on Telegram channels and Russian propaganda sites and was posted by Russian foreign ministry spokesperson Maria Zaharova.
Original report: Russia's Zaharova shares deepfake of Moldova's election chief
A federal judge in the Northern District of California ruled that Workday must identify customers who enabled its HiredScore AI features, and that a collective action alleging its applicant recommendation software disadvantaged candidates aged 40 and up includes applicants screened by them. Workday, which argued HiredScore is a separate product, must propose a timeline for identifying them by Aug. 20.
Original report: Judge orders Workday to identify employers using HiredScore AI
Users shared screenshots of xAI's Grok replying to a photo of Argentine streamer Milica (Mica Ybañez) with insulting, sexualized comments, at one point saying it would have sex with the streamer if it were human. The episode follows earlier criticism of Grok's pro-Nazi and antisemitic responses; Elon Musk has not commented.
Original report: Grok makes sexualized remarks about Argentine streamer Milica
Other people's shared ChatGPT conversations can be found on Google with a 'site:chatgpt.com/share' search, TechRadar reports. Chats go public only when users create a share link, but the prompt saying 'anyone with the URL' can view them does not make entirely clear that Google can index them, so a link meant for one person makes the whole chat public.
Original report: Shared ChatGPT conversations turn up in Google search
Scammers impersonating Keanu Reeves took $40,000 from one woman and close to $170,000 from another, according to the Marion County Sheriff's Office, using AI-generated voice and video to convince one victim that she was talking to the actor. Contact began through phone word games such as Scrabble, and the money was sent in Bitcoin to an account in Nigeria.
Original report: AI-faked Keanu Reeves romance scam costs two women
A Kuching-based influencer said in a July 30 statement that she lodged a police report after her livestream footage was allegedly taken without consent and altered with AI into a fake video endorsing an illegal online gambling website. Wynson Ong, founder of the International Legacy Education Foundation, said a complaint to the Malaysian Communications and Multimedia Commission will follow.
Original report: Kuching influencer reports AI deepfake promoting illegal gambling site
