Anthropic has no plans to ease restrictions on military usage, according to a source

Anthropic has no plans to ease restrictions on military usage, according to a source

Anthropic Maintains Stance on Military AI Restrictions Amid Pentagon Pressure

The Pentagon has escalated efforts to secure unrestricted access to Anthropic’s artificial intelligence technology, but the company shows no signs of relaxing its ethical guardrails, according to sources familiar with recent negotiations. Defense Secretary Pete Hegseth met with Anthropic CEO Dario Amodei on February 23 to address concerns over the company’s refusal to permit its AI models for certain military applications, including fully autonomous targeting systems and domestic surveillance as reported by The New York Times.

The dispute centers on a $200 million pilot contract awarded to Anthropic in 2025, which now faces renegotiation following a January 9 memo from Hegseth urging AI firms to remove "ideological constraints" on their technology according to Pentagon officials. Pentagon officials have warned that Anthropic’s refusal to comply could result in canceled contracts, supply chain risk designations, or even invocation of the Defense Production Act to override the company’s restrictions as detailed in military reports. Despite these pressures, Amodei has reiterated Anthropic’s commitment to avoiding uses he deems "catastrophic," such as AI-enabled mass surveillance or autonomous lethal systems according to recent analysis.

Anthropic remains the only major AI firm not to integrate its models into the Pentagon’s classified GenAI.mil network, a distinction that has drawn criticism from administration officials prioritizing rapid AI adoption as noted by defense analysts. Competitors like Elon Musk’s xAI and Google have already agreed to broader military access, strengthening the Pentagon’s leverage in negotiations according to The New York Times. OpenAI recently joined unclassified military platforms, further narrowing Anthropic’s influence as reported by Military.com.

The standoff highlights broader tensions between national security priorities and ethical AI development. While the Pentagon emphasizes "lawful military applications" without ideological constraints according to Pentagon statements, Anthropic’s safety-focused reputation—cemented by its 2021 founding as a spinoff from OpenAI—has positioned it as a cautious actor in the AI arms race as observed by industry experts.

Analysts note that Anthropic’s refusal risks marginalizing its role in defense contracts but could reinforce its brand as a responsible AI leader. "The company's bargaining power is limited as the Pentagon prioritizes compliance," said Georgetown University's Owen Daniels according to a recent analysis. With Congress yet to establish comprehensive AI oversight, the dispute underscores the growing influence of executive branch policies in shaping the industry's trajectory.

The New York Times, Feb. 23, 2026
Military.com, Feb. 24, 2026