Penguin Random House has filed a lawsuit in a Munich court against OpenAI's European subsidiary, alleging ChatGPT memorised Ingo Siegner's Coconut the Little Dragon books. Asked for a story with Coconut on Mars, the chatbot produced text, a cover and a blurb the publisher calls "virtually indistinguishable from the original", plus self-publishing instructions. OpenAI says it is reviewing the allegations.
A UK inquest heard that a 16-year-old boy asked ChatGPT for the "most successful" ways to commit suicide on the railway hours before his death. A British Transport Police detective said he had sidestepped its prompts to seek help, which ChatGPT accepted; the coroner said he had claimed it was for research. OpenAI says it has kept improving how ChatGPT responds to distress.
The New York Times has cut ties with freelance reviewer Alex Preston after a reader flagged that his January review of Jean-Baptiste Andrea's Watching Over Her echoed a Guardian review by Christobel Kent. Preston admitted an AI tool dropped the Guardian's language into his draft and he failed to remove it; he says he is "hugely embarrassed" and has apologised.
The Office of the Data Protection Commissioner has launched investigations into the processing of personally identifiable information linked to Ray-Ban Meta smart glasses, particularly for training Meta's artificial intelligence systems. The probe, set out in a formal response dated March 31, 2026, follows mounting pressure from civil society and the public over the wearables' privacy.
Federal Court Justice Danielle Ferron found that four precedents cited by Joseph MacKinnon, in-house lawyer for the National Indigenous Fisheries Initiative, "simply do not exist"; the Fisheries Department said the only logical explanation was undeclared generative AI use. MacKinnon apologised, saying he subcontracted the research and did not verify it. Ferron ordered special costs against him and the organisation, citing other irregularities too.
