Pentagon Flags Anthropic as Supply-Chain Risk
▼ Summary
– The U.S. Department of Defense has designated Anthropic as a supply-chain risk, a label typically reserved for foreign adversaries, due to a dispute over the company’s refusal to allow certain military AI uses.
– Anthropic’s CEO refused Pentagon demands to use its AI for mass surveillance of Americans or for fully autonomous weapons without human oversight in targeting or firing decisions.
– This designation threatens operations for both parties, as the U.S. military currently relies on Anthropic’s Claude AI in its Iran campaign and for data management via Palantir’s system.
– Critics, including former officials and hundreds of tech employees, call the move unprecedented and inappropriate, urging Congress to push back against treating a domestic innovator as a threat.
– In contrast to Anthropic’s stance, OpenAI has forged a deal with the Pentagon allowing military use for “all lawful purposes,” raising internal concerns about potential misuse.
The Department of Defense has formally identified Anthropic, a leading artificial intelligence company, as a supply-chain risk, a significant and unusual step typically applied to foreign entities. This designation stems from a fundamental disagreement over the ethical use of AI technology, specifically Anthropic’s refusal to permit its systems to be used for domestic mass surveillance or fully autonomous weaponry. The move creates substantial operational complications, requiring any Pentagon contractor to certify they do not use Anthropic’s models, despite the military’s current reliance on them in active campaigns.
The conflict centers on CEO Dario Amodei’s firm stance against allowing the military to deploy Anthropic’s AI for certain purposes. He has declined to let its systems power mass surveillance programs targeting Americans or to operate in lethal autonomous weapons where humans are removed from critical targeting and firing decisions. The Pentagon has countered that a private contractor should not be able to limit how the government utilizes artificial intelligence. This impasse led to the supply-chain risk label, which critics argue is an unprecedented and heavy-handed tactic against a domestic innovator.
The designation threatens to disrupt both Anthropic’s business and ongoing Pentagon operations, given the company’s unique position as the only frontier AI lab with systems ready for classified work. Reports indicate the U.S. military is actively using Anthropic’s Claude model within its Iran campaign, where it helps manage operational data at speed. Claude is a core component of Palantir’s Maven Smart System, a platform relied upon by military operators across the Middle East. The new risk status could jeopardize these critical integrations.
Several prominent voices have condemned the Pentagon’s action. Dean Ball, a former AI adviser in the Trump White House, described the designation as a “death rattle” of the American republic, accusing the government of abandoning strategic respect in favor of a “thuggish” approach that treats domestic companies more harshly than foreign rivals. Furthermore, hundreds of employees from firms like OpenAI and Google have publicly urged the Defense Department to withdraw the label. They have also called on Congress to intervene against what they see as an inappropriate exertion of authority, while encouraging industry leaders to maintain a united front against using AI for mass surveillance or autonomous killing.
In a contrasting development, OpenAI has secured its own agreement with the Pentagon, allowing military use of its AI for “all lawful purposes.” This broad phrasing has raised concerns among some of its staff, who worry it could open the door to the very applications Anthropic sought to prohibit. Amodei has characterized the Pentagon’s actions as “retaliatory and punitive,” suggesting his lack of political support for former President Trump contributed to the dispute. This stands in sharp relief to OpenAI president Greg Brockman, a staunch Trump supporter who recently made a substantial donation to a related Super PAC.
(Source: TechCrunch)
