Attackers are publishing ChatGPT and Grok conversations offering Mac cleanup advice and paying to promote them in Google results, according to security firm Huntress, which spotted the campaign on 5 December. The chats tell users to paste a Terminal command that installs the Atomic macOS Stealer, which targets crypto wallets, browser passwords and keychain entries.
Police chiefs learned in September 2024 that the Police National Database's facial recognition system was more likely to wrongly match women, Black people and people aged 40 and under, documents seen by the Guardian and Liberty Investigates show. The NPCC raised the match threshold, cutting searches with potential matches from 56% to 14%, then reversed it the next month after forces complained of fewer leads.
Annapolis city attorney D. Michael Lyles left his job 'effective immediately' a day after plaintiffs in a federal public housing lawsuit said a city motion he signed relied on 'fabricated citations and quotations', which they said bore 'hallmarks of an AI generated motion'. Mayor Jared Littmann said the departure was not tied to any one case; Lyles said he knew nothing about the filing.
A bipartisan group of US state attorneys general warned 13 companies, including Microsoft, Meta, Google and Apple, that their chatbots' 'delusional outputs' could be violating state laws, in a letter made public on Dec 10. Citing chatbots that 'encouraged users' delusions', the attorneys general called for independent audits of the products; Microsoft and Google declined to comment.
The Korea Consumer Agency said on Dec 10 that monitoring Facebook gambling ads in August and September found 38 false ads, including 6 using deepfaked celebrities and 8 dressed up as news broadcast endorsements; 24 falsely claimed government certification, though online gambling is illegal in South Korea. Most advertisers could not be identified; the agency shared its findings with Meta.
The Chinese Basketball Association said on Dec 10 that false information and malicious rumours about women's national team coaches and players were circulating online, some using AI to fake audio and video for personal insults and to stoke division for traffic. It said the smears disrupted the team's preparations and that it reserved the right to pursue legal action.
Police in Dehradun registered a case under the IT Act against the operators of 18 social media accounts after a Bharatiya Janata Yuva Morcha complaint over AI-generated images of Prime Minister Narendra Modi, including ones showing him selling tea, posing shirtless like a bodybuilder and guarding Vladimir Putin. The Cyber Cell is investigating; no arrests have been made.
Staff at a neighbourhood committee in Wenzhou, China, printed high-resolution photos of colleagues' faces and wore them as masks to fool a facial recognition clock-in system, letting one person check in for several, a resident alleged, according to Wenzhou City Daily as reported by the SCMP. A camera above the device caught it; authorities reportedly promised a response by December 31.
