Senior Trump administration officials met with Anthropic's Washington-based
executives on Tuesday to discuss AI safety risks, sources said. Anthropic chief
compute officer Tom Brown joined by video to speak with Pentagon CTO Emil
Michael and Commerce Secretary Lutnick. The meeting follows a 3,800-word essay
by Anthropic CEO Dario urging government regulation and a slowdown in
development of advanced AI systems so researchers can better assess potential
threats. Sources said Michael, who led Pentagon efforts to accelerate AI in the
military, has publicly rejected former Anthropic researcher Coxson's claim that
AI could kill all humanity; Lutnick is reportedly keen to rebut negative
messaging ahead of the November midterm elections. Anthropic declined to comment
and the White House did not respond.
