Anthropic: Is the Pentagon overreaching on AI access?
Deutsche Welle
February 25, 2026
AI-Generated Deep Dive Summary
The US Department of Defense is escalating tensions with AI startup Anthropic over access to its advanced models, raising concerns about the implications for national security and corporate autonomy. Defense Secretary Pete Hegseth has reportedly threatened to cut Anthropic off from Pentagon supply chains or invoke the Defense Production Act—a Cold War-era tool that allows the government to compel industries deemed critical to national defense—to force compliance. Anthropic, known for its focus on AI safety and ethics, has refused to grant unrestricted access to its Claude model, particularly for uses like autonomous military targeting or domestic surveillance. The company argues these applications pose significant risks given current safeguards.
The situation highlights a growing conflict between the Pentagon's desire to leverage cutting-edge AI for strategic advantage and Anthropic's commitment to responsible technology development. Since November 2024, Anthropic has been supplying Claude models to US intelligence and defense agencies, including during a high-profile operation in Venezuela that led to the capture of President Nicolas Maduro. While Anthropic has maintained a cautious stance on military applications, the Pentagon sees unrestricted access as essential for national security.
The potential invocation of the Defense Production Act marks an unprecedented move to control a private AI company, sparking fears it could stifle innovation and set a dangerous precedent for future tech governance. Critics warn that such measures might undermine Anthropic's ability to remain a leader in responsible AI development. As global competition in AI heats up, the balance between national security needs and ethical considerations will likely become a defining issue in technology policy.
Verticals
worldpolitics
Originally published on Deutsche Welle on 2/25/2026