Hamburg publisher Carlsen has sued OpenAI's Irish subsidiary, saying its rights and those of author Marc-Uwe Kling and illustrator Astrid Henn are infringed because simple ChatGPT prompts yield 'NEINhorn' stories, near-indistinguishable illustrations and even print-ready templates with a fake order number and publisher logo. Carlsen says the detailed copies indicate the books were unlawfully used to train OpenAI's models.
Denmark's Board on Research Misconduct unanimously found in May 2025 that University of Copenhagen sports historian Hans Bonde, at least through gross negligence, used AI to construct ten fictitious passages and references in a research article, its first such finding involving AI. Bonde admits errors but disputes the ruling, calling the material 'obviously fictitious' hallucinations; his complaint was rejected.
People-search service ClarityCheck, which assures users their image searches are 'private and secure', left more than 9 million image files, including photos of adults, teenagers and children, in an unsecured Amazon S3 bucket, according to researcher Jeremiah Fowler. The company secured it after WIRED's July inquiry but disputes that the data was 'publicly exposed' or held 9 million unique images.
A Washington Post investigation counted at least 69 police officials accused, charged or convicted of misusing Flock or other plate readers, and itself identified three employees in departments that did not regularly audit searches. One Indianapolis officer appeared to run 3,759 searches over 10 months on vehicles of his wife and acquaintances; he was suspended without pay and is under investigation.
Prosecutors charged former Kenosha County sheriff's deputy Frank McGrath with two felony counts of misconduct in public office, alleging he ran a romantic partner's plate in the Flock system 16 to 17 times while off duty and tracked her squad car and body camera about 40 times. Internal documents say he admitted the searches were jealousy-driven.
Fábio Luís Lula da Silva has sued presidential candidate Romeu Zema in São Paulo, seeking removal of 'Os Intocáveis', a satirical AI video published 10 July that depicts him in a marijuana plantation and has his lookalike call himself the 'Messi of Impunity'. His lawyers allege harm to his honour and image, seek R$10,000, and say the AI disclaimer is barely legible.
The NSA, FBI and other federal agencies warned of an 'active threat' in which unidentified attackers use AI-generated exploitation scripts, disguised as legitimate monitoring tools, against internet-exposed Siemens S7 controllers in US critical infrastructure. The advisory calls AI-assisted development an 'evolution' that slashes the expertise needed, and urges operators to isolate PLCs and patch.
An Amnesty International report says Javier Milei's government expanded AI-based surveillance, buying US$1.2 million of tools in 2024–2025 including social-media monitoring, facial recognition, drones and thermal cameras. Drawing on 21 interviews and procurement records, Amnesty says the tools have targeted dissent and deterred protest, with some journalists, activists and migrants self-censoring.
For a second straight day, hoax bomb threats disrupted schools in Wheatland, California, and investigators say the calls used an AI-generated voice. Bear River Middle School cancelled Wednesday classes and Wheatland Union High School started late while police, the FBI and other agencies swept both campuses with explosive-detection dogs; districts added overnight security and patrols.
A Bronx legal resident paid $4,820 via Zelle after answering a social media ad promising citizenship without the official exam, WABC reports; the ad was AI-generated from the likeness of a real Miami immigration attorney, who says he had no involvement. The attorney says close to or over 8,000 fake profile videos of him have been taken down.
