Pentagon Demands Unfettered AI for War: A Familiar Refusal of Restraint
Defense Secretary Pete Hegseth is reportedly summoning Anthropic CEO Dario Amodei to the Pentagon for discussions concerning the military's usage of Claude AI. The meeting is described as ‘tense’, with Pentagon officials expressing frustration over Anthropic's refusal to entirely lift safeguards on its technology. These safeguards specifically prevent mass surveillance of Americans and the
development of weapons capable of autonomous decision-making in targeting, areas Anthropic seeks to keep walled off. The military's reliance on Claude, cited as the most capable AI for sensitive defense and intelligence work within classified systems, underscores its strategic importance. The Pentagon, having helped fund foundational AI research, now demands unconditional access. This current
standoff echoes previous instances where military objectives clashed with ethical considerations in technological development. For example, during the Cold War, Project ARTICHOKE was launched by the CIA in 1951, exploring mind control techniques including hypnosis and drugs. This clandestine program, aimed at manipulating human behavior, highlighted a consistent willingness by elements within the
US establishment to push technological boundaries without adequate ethical oversight or public accountability. The same institution now expresses 'fury' that a private company refuses to deliver a product capable of independent lethality or domestic surveillance, despite the US military being responsible for an estimated 1.2 million direct and indirect deaths in its post-9/11 wars globally, a