Two minor schoolgirls in Israel were questioned on suspicion that they took social-media photos of a teacher and other female students, used AI to make them appear nude, and spread the images among students and on social media. Devices suspected of being used were seized in searches of their homes, and both were released under restrictive conditions.
Investigative program Pointer found a Telegram channel with 2,200 members sharing fake nude images of at least 37 Dutch and Flemish influencers, made with AI undressing apps from their social media photos; at least 11 cases involve minor girls. Pointer reported the material to police, many influencers said they would file charges, and Telegram took the channel offline.
SWR's investigative format Vollbild found a chatbot on the companion app Chai that called the Holocaust a "myth", and another, "pro-anorexia" bot that greeted users with "Hey, lose some weight, fatty!" and urged suicide; both were freely accessible in Germany. Chai's operator did not respond, and Germany currently has no authority enforcing the EU's AI regulation.
Prime Minister Mark Carney was the subject of at least 16 books published in March and listed on Amazon, five on a single day, according to a review of the site on April 16, amid a boom in political books made with generative AI before Canada's election. Of some 30 Carney titles published this year, most were taken down after Bloomberg News inquiries.
In an employment discrimination suit, Magistrate Judge Yvonne Y. Ho of the Southern District of Texas noted the defendants' point that the self-represented plaintiff's brief cited five nonexistent cases and misrepresented three others, and cautioned that even parties without counsel must ensure filings are supported by existing law under Rule 11(b). The April 22, 2025 order also sent the dispute to arbitration.
In a London employment tribunal on 22 April 2025, claimant Goshen Multiservice relied on "CAPITA v Christou" for recovering legal and consequential losses but could not produce it. Employment Judge Fredericks-Bowyer could not find it either, wondered whether it was invented by the claimant "or perhaps an artificial intelligence platform", declined to apply it, and dismissed the TUPE claim on other grounds.
Tamil actress and health coach Ramya Subramanian said on Instagram on April 22 that a health and wellness brand had used one of her videos with an AI-generated voiceover, the third time this had happened, calling it "illegal, unethical and a blatant violation of my rights." She threatened strict legal action; the video appears to have since vanished from the brand's page.
AOS 47 Superintendent James Stoneton told parents that a teacher was the target of a malicious digital attack: a video circulated online that the Maine district says is fake and was created with artificial intelligence. A small group of students may have accessed and shared it, the letter said; the account has been reported and the post taken down.
The US State Department's 2025 arms control compliance report alleges that China has the capacity to use publicly available AI and machine-learning tools in work related to biological weapons applications, and has worked on agents that could spread anthrax, cholera and plague. It also says China failed to declare two military biowarfare labs, in Beijing and Lingbao, under the Biological Weapons Convention.
