GOP rep terrified after ‘closed-door demo’ of new tech 'showed how to kidnap a lawmaker'

A prominent Republican House member was left scared after getting a first look at a new technology during a recent “closed-door demonstration,” Punchbowl News reported Friday — technology that the lawmaker warned most of their colleagues failed to comprehend the ramifications of.That technology was Claude Mythos, the latest large language model developed by Anthropic using generative artificial intelligence, the capabilities of which were so great that the Trump administration suspended its use by foreign nationals, effectively blocking its public release for the immediate future.“[Anthropic] told the model to find a vulnerability in a bank and empty accounts, and then it went and did it,” said Rep. Andrew Garbarino (R-NY), the chair of the House Homeland Security Committee, speaking with Punchbowl News. “[Mythos] then could find this vulnerability and fix it.”Beyond the Mythos’ apparent ability to drain bank accounts, Anthropic also demonstrated far more nefarious uses for its latest technology.“One unspecified model showed a detailed plan on how to kidnap a lawmaker in 30 seconds,” Punchbowl News’ report reads, paraphrasing Garbarino’s remarks. “Bad actors could hack the models and use them to attack critical infrastructure, [Garbarino] said, noting that AI’s agentic capabilities would allow models to take action without human prompting.”Citing specific examples he feared could be employed by those with access to Mythos, Garbarino warned that individuals could “turn off this gas pipeline” or “increase the amount of chlorine” in a community’s water supply, and all from an individual’s home.“You have the ability to not just learn how to do it, but to tell it to do it for you,” Garbarino told Punchbowl News. “I’d say 95 of my colleagues don’t understand what the hell’s going on.”While the Trump administration has, in effect, blocked Mythos’ public release for the immediate future, President Donald Trump has handicapped states’ ability to regulate the emerging AI technology with an executive order he signed last December that blocks states from creating their own laws to regulate the technology.