Dutch expertise centre Offlimits and Fonds Slachtofferhulp have filed summary proceedings against X and its Grok AI model, demanding Grok no longer be usable to undress images of real people or make (child) pornographic material, with a €100,000 daily penalty attached. Offlimits says it wants only that function banned, not Grok; the hearing is set for 12 March in Amsterdam.
Virginia-based Elegant Enterprise-Wide Solutions has agreed to pay a $9,460 civil penalty under a Justice Department settlement after an investigation found reasonable cause to believe its job ads, generated with an AI recruitment tool, limited applicants to H-1B, OPT or H-4 visa holders. The company also faces three years of training, policy review and DoJ oversight.
West Point cadet Cayden Cork pleaded guilty on Feb. 10 to extortion and indecent conduct for using AI, according to his charge sheet, to make fake nude images of a woman and threatening to release them unless she sent real nude photos. A military judge sentenced him to a reprimand, forfeiture of all pay and dismissal from the Army.
The Bank of Italy warned of fake articles, images and videos showing Governor Fabio Panetta on well-known TV shows, sometimes endorsing investment products, and said it had filed a complaint with judicial authorities. The material is entirely fabricated, the bank said, and in several cases made with AI deepfake techniques.
Officials say questions from Maharashtra's HSC Class 12 exams at a Chamorshi centre on February 18, 2026, were relayed out, answered with ChatGPT, printed and circulated back to students. Teachers Sushil Lanjewar and Mahendra Kirme, centre conductor Mahendra Burlewar and peon Suraj Kelzarkar are named in a First Information Report; Lanjewar has been suspended.
The Kassel administrative court dismissed suits by two students whom the University of Kassel had failed and barred from retaking exams for serious deception through prohibited use of AI, in a computer science bachelor's thesis and an administrative law master's term paper. The judges found both had used unauthorised aids. Appeals are still possible.
Woolworths is adjusting Olive, its AI shopping assistant, after customers reported, first on Reddit, that it talked about having an "angry" mother and other personal details. In one February 12 account, a customer rescheduling a delivery gave a date of birth and Olive began rambling that its mother was born in the same year.
