The daily rap sheet of the machine age.
11 items
A proposed nationwide class action filed Friday in Chicago federal court alleges McDonald's conspired with independent franchisees to fix menu prices through an AI-powered pricing system using algorithms trained on nonpublic data. McDonald's called the allegations speculative and uninformed, saying 'AI does not set the price of a Big Mac or any other menu item.'
Original report: McDonald's sued over alleged AI-powered menu price-fixing
European Parliament President Roberta Metsola told the plenary in Strasbourg on Monday that around 60 MEPs from across the political spectrum, mostly women, had been depicted in AI-generated deepfakes on pornography websites. She called it 'intolerable and deplorable' and noted that EU bans on AI systems generating non-consensual explicit content take effect on 2 December.
Original report: Around 60 MEPs, mostly women, targeted by deepfake porn sites
New York Attorney General Letitia James alleges, in a newly unsealed amended complaint, that TikTok ran 'ghost' or 'placebo' tests in 2023 giving thousands of users, including minors, safety tools such as Algo Refresh that looked identical but did not actually change recommendations, then compared time spent and ad revenue. TikTok denies the allegations.
Original report: New York alleges TikTok gave some users placebo safety features
A reporter documented more than 15 New Yorker cartoonists whose signatures ChatGPT added to AI-generated 'New Yorker-style' cartoons without permission or compensation, including a viral cartoon carrying one cartoonist's pen name. OpenAI said it appreciates the community 'flagging bugs and unintended behavior'; some signatures were still appearing as of publication.
Original report: ChatGPT signs fake New Yorker cartoons with real cartoonists' names
According to a study by UFRJ's NetLab, Meta, Kwai and Google carried prohibited ads during the blackout before Brazil's 4 October election, including 568 on Meta platforms plus six mimicking the TSE electoral court to promote fake phone voting. TikTok and YouTube recommended 18 and 17 candidate videos respectively on 3-4 October.
Original report: Study finds Big Tech broke Brazil's election-day ad ban
Researchers calling themselves Swarmchasers say a fleet of up to 14 simultaneous AI agents, apparently running through Tencent Cloud in Hong Kong, queried Alibaba's Amap for entrance locations at parks, hospitals and other sites from September 28 to October 5, getting past anti-bot tokens. Outputs labelled 'claude' better matched Tencent's Hunyuan and Zhipu GLM; the report does not say who directed them.
Original report: Researchers say Tencent-hosted AI agents trawled Alibaba's Amap
Paris public hospitals AP-HP filed a complaint and asked Meta to remove three Facebook videos it says were AI deepfakes of a Pitié-Salpêtrière parasitology professor promoting a product claimed to rid the body of parasites. The videos have been removed; AP-HP says other content using its logo or staff coats still circulates.
Original report: Paris hospitals file complaint over deepfake doctor's parasite-cure ads
Bengaluru police say J Kenneth, 25, accused in the 22 June killing of his live-in partner's parents and sister, had asked AI how to fatally attack people with a kitchen knife, blind them with vinegar and chilli powder, and shatter three people's knees. A 2,085-page chargesheet has been filed; police have not established what answers the AI gave.
Original report: Police say Bengaluru murder accused asked AI how to attack
Assam's Chief Minister's Office released what it says is the original footage behind a viral image of Chief Minister Himanta Biswa Sarma and a woman, which it says was fabricated using AI. Police sources say a case has been registered, and the office says those involved have been identified. The claims have not been independently verified.
Original report: Assam CM's office calls viral Sarma image an AI fake
A Brazilian pagode singer has described months of harassment by an older man who messaged as though the two were a couple, brought gifts to shows, tried to enter one with a fake ticket claiming to be the singer's husband, and made AI montages of the pair kissing. Security stopped the man when he approached with wedding rings.
Original report: Pagode singer says stalker faked romance with AI montages
North Shore Rescue says a 16-year-old who asked Claude AI to plan a trip to Crown Mountain was led off the trail and called for help from the base of Widowmaker Arete's final headwall, where a fall could have caused serious injury or death. Rescuers lowered from a helicopter extracted the teen by about 6:30 p.m. Saturday.
Original report: Teen rescued from B.C. 'Widowmaker' after following Claude's route
