The daily rap sheet of the machine age.
8 items
Colombian senator Carlos Mario Farelo Daza says that since April 2025 scammers using the senator's photo, official logos and an AI-cloned voice have offered fictitious Congress jobs for money, defrauding at least 14 people of more than 15 million pesos. Farelo has filed a complaint with the Fiscalía General de la Nación and says it is acting too slowly.
Original report: Cloned voice of Senator Farelo used to sell fake jobs
An appeals court in Zárate-Campana unanimously ruled that Article 128 of Argentina's Penal Code covers representations of child sexual abuse generated or modified with AI, even without identified real victims. Rejecting a defence that no real victims were proven and the content might be digitally created, it upheld sending to trial a man accused of publishing, distributing and possessing such material.
Original report: Argentine appeals court says AI-generated child abuse imagery is criminal
A Civio investigation reports that Quantus Skin, a melanoma-screening algorithm from Asisa-owned Transmural Biotech that Basque health service Osakidetza is testing under a €1.6 million contract, showed 69.1% sensitivity in a published study, missing one in three melanomas and below the contract's 85% threshold. It was trained almost exclusively on images of white patients; Transmural's CEO blamed the weak results on poorly captured images.
Original report: Basque skin-cancer AI misses one in three melanomas, study finds
Judge Elia Nos of the Petah Tikva Family Court found that none of the rulings or scholarly works an applicant's lawyer cited in an estate dispute could be found, invoking Supreme Court criticism of filings resting on nonexistent sources, some apparently AI-assisted. The court rejected the application and ordered the applicant to pay increased costs of ₪3,000 and ₪4,000.
Original report: Court's search for a lawyer's cited rulings comes up empty
A satirical TikTok video, which detection tools Hive Moderation and Hiya flagged as likely AI-generated, shows a figure simulating Brazil's First Lady Janja da Silva boasting of Paris luxury shopping with taxpayers' money; Projeto Comprova found no evidence for the claims. The video drew 160,000 views and some commenters believed it. Official data put the whole delegation's trip at R$60,210.58.
Original report: TikTok fake shows AI Janja flaunting taxpayer-funded luxury
An elderly couple travelled more than 300km from Kuala Lumpur to Perak to ride the "Kuak SkyRide", a cable car shown in a nearly three-minute AI-generated video featuring a "TV Rakyat" reporter, before a hotel worker told them it did not exist. Police say preliminary checks confirmed the video was entirely fabricated and no such project exists.
Original report: Elderly couple travel over 300km to ride AI-invented cable car
In an employment suit against the Oklahoma Medical Marijuana Authority, US District Judge Scott L. Palk found citations in the plaintiff's response brief inaccurate and nonexistent, with quotes that cannot be located, signalling counsel may have used AI "or otherwise counsel produced exceptionally sloppy work". The court issued a strong admonishment, warning counsel of possible sanctions from the state court on remand.
Original report: Judge says lawyer's fake citations suggest AI, or exceptional sloppiness
US District Judge Douglas R. Cole ordered a pro se plaintiff suing Gap to show cause by July 11, 2025 why he should not be sanctioned for citing cases that do not appear to exist and quotes the court cannot find, which the court suspects came from generative AI. Possible sanctions include dismissal; a hearing is set for July 16.
Original report: Plaintiff suing Gap ordered to explain apparently invented cases
