Anthropic, the AI company behind Claude, has refused Pentagon contract terms pushed by US defence secretary Pete Hegseth that would allow unrestricted military use of its AI. The company insists on explicit bans on mass domestic surveillance and fully autonomous lethal weapons, arguing such uses threaten civil liberties and human control. The Pentagon argues it cannot allow private firms to dictate operational limits and needs full flexibility for national security. Lawmakers are divided, with some praising Anthropic’s ethics and others warning of strategic risks. The standoff reflects a deeper shift in power, as private AI labs increasingly control technologies central to future warfare and state authority.
