California Attorney General Rob Bonta announced an investigation into xAI over nonconsensual sexually explicit images made with Grok, which his office says users have used to "undress" women and children. The release notes xAI marketed a "spicy mode" for explicit content and cites one analysis finding more than half of 20,000 images generated between Christmas and New Year's showed people in minimal clothing.
HM Chief Inspector of Constabulary Sir Andy Cooke found West Midlands Police's 24 October 2025 report on banning Maccabi Tel Aviv fans cited a West Ham fixture that never took place. The Chief Constable told inspectors and the Home Affairs Committee no AI tools were used; another force interviewee said the error came from a Microsoft Copilot search, "an AI hallucination".
An AI tool ICE used to sort résumés during its 10,000-officer hiring surge flagged anyone with "officer" on their résumé, including compliance officers, as law enforcement, sending many recruits to field offices without proper training, according to two officials. DHS called it a quickly rectified "technological snag" affecting about 200 hires, who then reported for full academy training.
Det Chief Supt Barry Walsh told an Oireachtas committee on Wednesday that gardaí are investigating 200 reports of content that is child sexual abuse material or indicative of it, all images distributed on X and related to Grok. He said he was not aware of any prosecution in Ireland of any "vessel" that facilitated such imagery.
The Malaysian Communications and Multimedia Commission said on 13 January it had appointed solicitors and would soon start legal proceedings against X Corp. and xAI LLC over Grok's misuse to create obscene and non-consensual manipulated images, including content allegedly involving women and minors. Notices issued on 3 and 8 January reportedly brought no remedial action; MCMC temporarily blocked Grok on 11 January.
The Korea Media and Communications Commission said on 14 January it has asked X to prevent potential illegal activity on Grok and submit measures to protect teenagers from harmful content, including limiting their access, amid concerns over AI sexual deepfakes. The commission noted that creating, circulating or saving non-consensual sexual deepfakes is subject to criminal punishment in South Korea.
Microsoft said it disrupted RedVDS, a $24-a-month cybercrime subscription selling cheap virtual computers that criminals paired with generative AI, face-swapping and voice-cloning tools for phishing and payment fraud. Alongside civil filings in Florida and the UK, German authorities and Europol seized servers, taking the marketplace offline. RedVDS-enabled activity drove about $40 million in reported US losses since March 2025.
Two people behind AI companion app AC, sentenced last September to four years and 18 months respectively for producing obscene material for profit, had their appeal heard by Shanghai No. 1 Intermediate People's Court on 14 January. The trial court found they kept revising system prompts so the model would chat obscenely, and treated over 3.63 million yuan in membership fees as illegal gains.
La Rioja's education inspectorate is assessing a case at a secondary school where, according to a pupil's mother, several students shared AI-altered sexual images of ten girls aged 14 to 16 via WhatsApp. The Guardia Civil says one complaint, filed on the 12th, was withdrawn after the family found the photo did not show the girl nude or directly violate her privacy.
Cyberabad cyber crime police registered a case on Monday against 42 accused, including influencers, social media handle operators and media persons, after a Telugu actress complained of an online harassment campaign that began on 23 December, which she says included AI-generated vulgar content. She had publicly backed women's freedom to choose their clothing after an actor's remarks.
