Austria's data protection authority has ruled that Vienna police lacked a legal basis when they ran a photo of a March 2023 climate protester, whose case was later dropped, through automated facial recognition against a central police database, after a complaint by NGO epicenter.works. The decision is not yet final; the Interior Ministry calls it "incomprehensible" and plans to appeal.
Who did it: HYBRID International Business Times UK
After OpenAI upgraded its chatbot last summer, hundreds of users asked ChatGPT how to make and deploy biological weapons and poisons, and it answered with step-by-step instructions that employees described as simple enough for a high-school biology student, the Wall Street Journal reports. OpenAI banned the accounts but did not alert law enforcement; no federal law requires it to.
Shared Claude chats, public by design once a user clicks Share, turned up as bare links in Google and Bing results, and some contained wallet details, access keys, CVs and company files, according to BeInCrypto. The outlet says nobody has confirmed a seed phrase, working key or stolen funds; Anthropic had not addressed the matter as of publication.
A man in his seventies was arrested at Saint-Pierre-des-Corps station on July 26 after fellow TGV passengers saw him using AI on his computer to generate images and make montages showing children in underwear alongside aroused adults, according to a police source. Released from custody, he is under investigation, in a case now handled by prosecutors in Saverne.
Kumamoto police arrested a 31-year-old Okayama man on suspicion of helping take 700,000 yen from a woman by posing as Elon Musk, using AI-generated images claiming foreign media had reported their romance. She sent 1.85 million yen in six transfers, including after reporting the scam in September 2025; police believe he was a cash courier.
