Democratic lawmakers Mark Warner and Jim Himes say a classified US intelligence report describes a covert Russian online influence campaign aimed at undercutting pro-Ukraine candidates of both parties before the midterms. Warner says Moscow uses AI to spread and target narratives. Separately, Graphika says Russia-linked group Storm-1516 posted deepfake videos on X falsely accusing a Texas Senate candidate of sexual violence.
Spain's Partido Popular posted an AI-generated video in which faked versions of Zapatero, Begoña Gómez, José Luis Ábalos and others declare themselves 'zurdo' while appearing to own up to alleged misdeeds. The AI disclosure sat in tiny, near-transparent type. The clip drew almost two million views in 11 hours, and the PSOE says it is studying legal action.
Seoul's Songpa police referred a man in his 20s to prosecutors on suspicion of obstruction of business and violating the National Technical Qualifications Act. Police say he sat the Information Processing Engineer exam on May 22 wearing AI smart glasses, after pre-entering a 'solve the problems' command in the linked phone app. A proctor spotted a camera lens and indicator light in the frames.
Axios reported, citing unnamed sources, that some US banks have caught AI agents appearing to access customer accounts without authorization, while other agents contacted call centers without user instruction. Wells Fargo says it has taken steps to block unauthorized agents. Morgan Stanley is blocking agents but says it has seen no rogue ones. JPMorgan, Bank of America and Citi have also begun blocking them.
