Essex Police paused live facial recognition after a University of Cambridge study, with 188 actors walking past cameras in Chelmsford, found the system statistically significantly more likely to correctly identify black participants than others, and more likely to identify men than women. The force says it has since revised its policies after working with the software provider and is confident it can resume deployments.
The Nebraska Supreme Court struck Omaha divorce attorney Greg Lake's appellate brief, which it said contained fictitious, potentially AI-generated cases, holdings and quotations, and referred him for discipline. Opposing counsel said 57 of its 63 references were defective. Lake blamed "sloppy" copying and pasting and a computer broken on vacation; the court affirmed the divorce decree.
Who did it: HYBRID AI Hallucination Cases database
Ohio's Eleventh District Court of Appeals sanctioned attorney William B. Norman after his paralegal used ChatGPT on a filing quoting inflammatory prosecutor remarks that were never made. It removed him as counsel, referred him for discipline and ordered apology letters. The court noted that his remedial AI policy itself looked AI-generated, "[Insert Date]" placeholders included.
Spain's National Police in Albacete arrested a man suspected of using AI to turn an acquaintance's photo into a nude image, which he allegedly sent to her on social media. Police say the victim is a minor and that, after she reported it, he confronted her in the street and threatened her mother's and relatives' lives unless she withdrew the complaint.
A woman is suing the Nova Scotia man who made nude deepfakes of five former high school classmates. She alleges invasion of privacy and cyberbullying over an image built from a photo taken when she was a minor. A judge ruled the deepfakes fell outside the Criminal Code's intimate-image definition; he pleaded guilty to criminal harassment and sending an obscene picture.
