Korea's Literature Translation Institute says 21 English-language books by Korean authors appear on the compensation list in the US class action accusing Anthropic of training Claude on illegally downloaded books, and has notified the authors. Yonhap reports the case closed in July with a $1.5 billion settlement, roughly $3,000 per work, payable to authors who file claims.
A 39-year-old Washington state man has sued OpenAI in a San Francisco civil court, alleging ChatGPT advised him on medication choices and doses, told him there were "no serious interactions" between hydroxyzine and Adderall, and backed his decision to stop a prescribed mood stabiliser. The complaint says this amounted to unlicensed medical practice and contributed to his involuntary psychiatric hospitalisation in 2025.
OpenCode says an intruder accessed a legacy system it was decommissioning and that stored third-party API keys, including OpenAI and Anthropic ones, may have been exposed for 1.4% of Console users; it links the breach to a large-scale distillation attack. Reddit users report unauthorised charges, including $400 on one Anthropic account and about €200 in Google usage.
An image purporting to show an elderly couple assaulted and stripped in Paris after the 17 October protests, viewed six million times in one X post and briefly shared by police officers' union SCSI, was AI-generated, the Observers newsroom reports. Google's SynthID detected OpenAI generation traces; the image was traced to a montage posted by pro-Israel site Infos-Israel.news.
An Indian actor posted on X on Saturday ordering anyone making AI-generated videos or images of her face or likeness without consent to stop and take them down, saying such use to misrepresent her or malign her reputation is unacceptable. She warned she would take legal action against everyone creating, posting or circulating the content, fake accounts notwithstanding.
Yonsei University caught an applicant who swapped into camera-equipped smart glasses after pre-exam checks during a 10 October maths essay exam, and the University of Seoul ejected another on 3 October, barring them from its entrance exams for three years. A newspaper reporter's test found such glasses displayed AI-generated answers in about a second; the education ministry is issuing proctor guidance.
