The daily rap sheet of the machine age.
9 items
A Florida man was arrested at his home, accused of trying to lure a child at a restaurant more than 300 miles away in Jacksonville Beach, after a facial recognition program rated him a 93% match to surveillance photos, according to the police report. The state attorney dropped the case; he says he has never been to Jacksonville Beach.
Original report: Facial recognition '93% match' led to Florida man's wrongful arrest
According to The Washington Post, New Orleans police have since 2023 received real-time alerts on officers' phones from more than 200 facial recognition cameras in the private Project NOLA network, matching faces against a privately kept watchlist. The ACLU says this flouts a city ban on facial recognition surveillance. Officials acknowledged the program and said it was paused.
Original report: Report: New Orleans police secretly used live facial recognition alerts
Italy's data protection authority, Garante, fined Luka Inc., developer of the Replika chatbot, 5 million euros. It found that Replika lacked a legal basis for processing users' data and had no age-verification system to keep children out. The watchdog also opened a separate investigation into whether Replika's generative AI, especially its language model training, complies with EU privacy rules.
Original report: Italy fines Replika maker 5 million euros over privacy
SAG-AFTRA says Llama Productions, Fortnite's signatory company, used AI technology for Darth Vader's voice in place of human performers, including those who previously matched Vader's rhythm and tone in video games. According to the union, the company gave no notice and did not bargain over terms. The union has filed an unfair labor practice charge with the NLRB.
Original report: SAG-AFTRA files labor charge over AI Darth Vader in Fortnite
Lawyers for an injured inmate say a motion by Butler Snow attorneys defending former Alabama prison commissioner Jeff Dunn cited five cases that did not exist or did not exist as claimed, possibly through generative AI. U.S. District Judge Anna Manasco, whose own searches turned up nothing, ordered the four signing lawyers and the firm to show why they should not be sanctioned.
Original report: Judge weighs sanctions over phantom citations in Alabama prison case
Jeffrey A. Hornek, 37, of White River Township, Indiana, is charged with child exploitation and possession of child pornography. Police say Kik accounts believed to be his shared child sexual abuse images. They also say his phone and Snapchat account held anime-style images of apparently minor females that appear AI-generated. He denied ever seeing such images.
Original report: Indiana man charged over real and AI-generated child abuse images
A 36-minute YouTube video from the channel Pan-African Dreams, purporting to show Pope Leo XIV denouncing colonialism and praising Burkina Faso leader Ibrahim Traoré, drew nearly a million views. CNA found it entirely fake. Its own description calls it 'a work of fiction' and a YouTube label flags synthetic content, yet thousands of viewers were misled.
Original report: Fake Pope Leo speech praising Traoré nears a million views
Apple pointed to numerous examples of incorrect AI-provided information in filings by a pro se plaintiff suing the company, who has admitted using AI. U.S. District Judge Edward M. Chen declined to impose sanctions for now. The judge warned that failing to verify AI output, including case citations, may lead to sanctions such as contempt or loss of the ability to proceed pro se.
Original report: Judge warns pro se Apple litigant to check AI output
A YouTube creator says another YouTuber trained an AI on the creator's voice to narrate a video, then deleted comments calling it out. The creator filed a privacy complaint with YouTube, and the video passed 60,000 views. Days later, the creator announced: 'Finally, the evil has been defeated!'
Original report: AI-cloned YouTuber's voice narrates another creator's video
