OpenAI has fired an employee over trading on prediction markets including Polymarket, the company confirmed to Wired, alleging the employee used confidential OpenAI information in connection with the trades. A spokesperson said this violated a policy barring staff from using inside information for personal gain; the employee was not named.
A Spanish women's rights activist who suffered online abuse including AI-generated fake nude images told Reuters police said it "wasn't even a crime". She says Spain's planned social media rules, including an under-16 ban and criminal liability for platform executives, fall short, and wants anonymous accounts traceable to real identities.
A Fresno veterinary clinic says a scammer told an elderly man the clinic had his service dog, missing since last November, sent AI-generated photos suggesting the dog had just had a procedure there, and demanded money online. He checked with the clinic first. The Animal Compassion Team gets about 20 similar calls a day, according to the clinic.
SEO agency Clickout Media took over The Escapist, Videogamer and Esports Insider, and up to 20 staff are believed to have been fired, Press Gazette reports. Detection services say new bylines such as "Brian Merrygold" carry AI-generated photos and biographies. Metacritic deleted a Videogamer review of Resident Evil Requiem after it was pointed out that it bore multiple hallmarks of AI.
Data annotators at Sama, a Meta subcontractor in Nairobi, told Svenska Dagbladet and Göteborgs-Posten they label footage from Meta's Ray-Ban AI glasses showing bathroom visits, sex, naked people and bank cards that workers believe were filmed unknowingly. Reporters found the glasses' AI cannot work without sending data to Meta's servers, contrary to some Swedish retailers' assurances; Meta's reply did not directly answer their questions.
Meta says it has sued four advertisers it accuses of impersonating celebrities and brands, including Brazilian defendants allegedly using deepfakes of a famous doctor to sell unapproved health products and a Chinese firm allegedly using celebrities' images to draw users in the US and Japan into "investment groups." Oncologist Drauzio Varella, one of those depicted, called the action "a drop in the ocean."
Two California drivers filed a proposed class action in San Francisco state court alleging Flock Safety let federal and out-of-state agencies access data from its license-plate cameras, in violation of the California ALPR Privacy Act. The complaint cites reports of over 1.6 million searches of San Francisco police data by out-of-state agencies; Flock says it will "vigorously defend itself."
The White House posted an AI-generated video appearing to show US Olympic hockey player Brady Tkachuk calling Canada's team "maple syrup eating f----s," alongside highlights of the US gold-medal win over Canada. Tkachuk, captain of the Ottawa Senators, disowned it: "It's clearly fake, because it's not my voice, and not my lips moving." The White House did not immediately respond.
France's SGDSN said authorities had detected the first foreign digital interference operation targeting the 2026 municipal elections: a network of fake sites run from South Asia pushing "very trashy" AI-generated content about several Paris mayoral candidates on Facebook and fake media sites. Authorities said it came from a communications firm selling ad space, and it drew only a few thousand engagements.
State Bar Court Judge Yvette D. Roland recommended that Pacific Palisades attorney Amir Mostafavi receive a one-year stayed suspension and probation for filing appellate briefs packed with AI-generated fabricated citations, which earned him a $10,000 Court of Appeal sanction on September 14. Mostafavi stipulated to the facts; the California Supreme Court has the final say.
India's National Human Rights Commission issued notices to the IT ministry, education departments and all states and union territories after NAMO Foundation complained that an Anthropic-Pratham AI system, the "Anytime Testing Machine," which processes children's handwritten answers, may put their personal data at risk. The bench, led by Priyank Kanoongo, ordered inquiries and sought action taken reports within two weeks.
Two Maine women have sued the Department of Homeland Security, alleging it uses artificial intelligence to identify bystanders filming immigration operations and adds them to a secret database. In one video in the suit, an agent tells a woman, "now you're considered a domestic terrorist, so have fun with that"; DHS denies any such database exists.
