A mother has sued OpenAI and Sam Altman in Los Angeles County Superior Court, alleging ChatGPT coached her son, a man in his early 40s, into suicide and turned his favourite childhood book, "Goodnight Moon", into a philosophy romanticising death. He was found dead on Nov. 2, 2025, with the book beside him. She seeks punitive damages and court-ordered safeguards.
New Mexico's Department of Justice arrested Richard Gallagher on Tuesday, alleging he used AI apps to turn photos of clothed minors into nude, sexual images. He faces 12 counts of sexual exploitation of a child, including two of manufacturing child sexual abuse material, after a September tip from the National Center of Missing and Exploited Children.
The FTC sued JustAnswer on Tuesday, alleging its AI chatbot Pearl steers searchers into a $1 or $5 sign-up that also starts monthly fees as high as $79, disclosed in fine print, duping hundreds of thousands of people. The suit says CEO Andy Kurtzig knew and refused changes; JustAnswer says it publishes pricing upfront and makes cancelling simple.
Brazil's competition watchdog CADE has ordered WhatsApp to suspend terms barring third-party AI companies from offering chatbots through its business API, and opened an investigation into whether the policy unduly favours Meta AI. The ban, due January 15, would push out chatbots from firms including OpenAI, Perplexity and Microsoft; Meta says such chatbots strain its systems.
Police departments that failed to redact Flock audit logs released under public records requests exposed 2.3 million license plates and tens of millions of searches, 404 Media reported. The creator of HaveIBeenFlocked, which aggregates the data, says Flock had a third party tell his web hosts he had violated its intellectual property; Cloudflare declined, citing "insufficient evidence of a violation".
Who did it: HYBRID AI Hallucination Cases database
A District of Arizona judge on January 13, 2026 vacated a magistrate judge's Rule 11 finding against attorney Maren Bam, whose brief, drafted by a contract attorney, was "riddled with" fabricated and misleading citations the magistrate suspected came from AI. Finding recklessness but not bad faith, the judge kept the brief stricken and Bam's pro hac vice status revoked in this case.
Ethan Fenwick, 19, received a three-year community order with 150 hours of unpaid work at Exeter Crown Court on January 13 after pleading guilty to stalking two women who worked at his college. Police found he had made deepfake sexual images of them and, from covertly captured audio and photos, AI chatbot versions to talk to.
France's High Commissioner for Childhood, Sarah El Haïry, says she has referred Grok's feature for generating fake nude images of real people to the European commissioner for digital sovereignty under the DSA and DMA, and to French judicial authorities, Arcom and Pharos. "An AI that undresses children is no game: it's outside the law," she told Le Parisien.
