The Pentagon has labeled artificial intelligence company Anthropic a “supply-chain risk” after its CEO, Dario Amodei, refused to grant the Department of Defense unrestricted access to its AI systems, a move that underscores a growing conflict between Silicon Valley and Washington over the control of advanced technology.
The designation, typically reserved for foreign firms considered potential adversaries like Huawei, stems from Anthropic’s refusal to allow the Pentagon to utilize its AI for mass surveillance or to power fully autonomous weapons systems, according to statements released by both parties. Amodei maintains that such applications would violate the company’s ethical principles, which prioritize AI safety.
Palmer Luckey, founder of defense technology company Anduril Industries, argues that the decision of how AI is deployed should rest with the government, not private companies. “We demand to stick to a position that this is in the hands of the people,” Luckey said in a recent interview with the Latest York Post. “Anyone who says that a defense company should be going beyond the law… you are effectively saying you do not believe in this democratic experiment, that you want a ‘corporatocracy.’”
Luckey’s position reflects a broader concern that private sector decisions could unduly influence U.S. Foreign policy. He contends that tech companies should adhere to the foreign policy directives of the administration in power, regardless of their own internal ethical considerations. “In all cases, whoever the United States government tells me that I can and cannot sell to,” Luckey stated, “to have any other position is to fall further into… basically corporate executives having de facto control over U.S. Foreign policy.”
The standoff with Anthropic follows a similar instance in 2018 when Google terminated its involvement in Project Maven, an AI program designed to analyze drone footage for the Pentagon, following internal employee protests over the potential for the technology to be used in autonomous weapons. This earlier incident, as noted by Luckey, highlights a recurring tension between tech companies and the defense establishment.
Although Anthropic has resisted the Pentagon’s requests, other companies have reached agreements to provide AI capabilities. OpenAI, led by Sam Altman, has agreed to allow the Pentagon to utilize its AI models and tools, as has Elon Musk’s xAI, creating a competitive landscape where different companies take divergent approaches to government partnerships.
Despite the disagreement, Amodei indicated Thursday that Anthropic and the Department of Defense are seeking common ground. “Anthropic has much more in common with the Department of War than we have differences,” he said in a press release, suggesting ongoing negotiations. Anthropic has indicated it will sue to overturn the “supply-chain risk” designation, but remains in discussions with the Pentagon regarding potential uses of its AI models.