Prosecutors say Daniel Moreno-Gama threw a lit Molotov cocktail at OpenAI CEO Sam Altman's home trying to kill him, motivated by hatred of AI. He faces an attempted murder charge and federal explosives and firearm charges; a federal complaint says a document he carried warned of humanity's "impending extinction" and listed the names and addresses of other AI executives, board members and investors.
Bern prosecutors have opened an insult and defamation investigation after Swiss Finance Minister Karin Keller-Sutter filed a criminal complaint against "unknown persons"; Sonntagsblick had reported that an X user allegedly prompted the Grok chatbot to hurl sexist insults at her. The European Commission has separately opened proceedings against Elon Musk's group following a controversy over sexualised images generated by Grok.
Ten minors aged 14 and 15 pleaded innocent at a closed-door juvenile trial in Valladolid over alleged AI-faked nude images of five female schoolmates in November 2024. Prosecutors, alleging child pornography and offences against moral integrity, seek twelve months of socio-educational tasks each; the defendants named one of them as the bot's creator, which he did not admit.
According to 404 Media, WebinarTV, which claims to host more than 200,000 scraped Zoom webinars, posted anonymous addiction recovery meetings, chronic illness support groups and a nudist gathering, some showing participants' full names and faces, and earlier turned one meeting into an AI-generated podcast. WebinarTV's Michael Robertson says webinars are public broadcasts and hosts are asked for permission.
SBT's "Se Liga Brasil" spent about five minutes on a supposed story, presented by Thiago Gardinali, about a woman calling military police over a misogynistic sign at a gas station; the image was AI-generated, and the presenter wrongly said the case reached a police station. SBT says airing it unchecked broke its journalism standards and internal measures were taken.
Yantai police in Shandong have closed more than 8,000 accounts after cracking two 'water army' cases in which groups used AI to mass-produce negative articles about new energy vehicles. Police say one gang rewrote netizens' posts, including one about a Li Auto car, to earn platform traffic fees, making up to 200,000-300,000 yuan a month.
Israel's High Court let media name brothers Meir and Yosef Nahum, indicted for espionage on behalf of Iranian agents to whom they allegedly passed information for tens of thousands of shekels. The indictment says Meir, the main defendant, fed the handler false material made with AI tools such as ChatGPT, Grok and Gemini, including an AI-generated photo of a fictitious Unit 8200 soldier.
Who did it: HYBRID AI Hallucination Cases database
Ashdod Magistrate Court ordered plaintiff Mutzarei Hashmal Hatzi Hinam to pay NIS 5,000 to the state treasury after its responses in a suit against Be'er Ya'akov municipality cited nonexistent regulations. The company admitted using AI only when asked why costs should not be imposed; the judge declined to charge its lawyer, who acted as its agent.
