The Pentagon has officially designated artificial intelligence company Anthropic and its products as a supply chain risk, effective immediately. This unprecedented move could compel government contractors to cease using Anthropic's AI chatbot, Claude.
The decision follows accusations from President Trump and Defense Secretary Pete Hegseth that Anthropic's products could endanger national security. The company's CEO, Dario Amodei, stated the company intends to challenge the designation in court, asserting it is not legally sound.
The Pentagon cited the principle of allowing the military to use technology for lawful purposes, asserting it will not permit a vendor to restrict critical capabilities. Amodei countered that Anthropic sought limitations on high-level usage areas concerning surveillance and autonomous weapons, not operational decision-making. He noted "productive conversations" had occurred regarding Claude's continued use or a smooth transition.
Some contractors, like Lockheed Martin, indicated they would comply with the directive and explore alternative large language model providers, expecting minimal impact due to not relying on a single vendor.
Microsoft stated its legal review suggests they can continue non-defense related projects with Anthropic.
The Pentagon's application of a rule typically for foreign adversaries drew criticism. Senator Kirsten Gillibrand called it a "dangerous misuse of a tool" intended for adversary-controlled technology. Former officials and policy experts also expressed "serious concern," viewing the designation against a domestic company as a "profound departure" from its intended purpose and a "category error."
Despite losing defense contracts, Anthropic has seen a surge in consumer downloads, with over a million daily sign-ups for Claude, topping AI app charts in numerous countries. This dispute also highlights Anthropic's rivalry with OpenAI, which announced a deal to supply ChatGPT for classified military environments shortly after the Pentagon's announcement.