Anthropic Challenges Pentagon's Supply Chain Risk Designation Amid AI Dispute

An unexpected conflict has erupted between the US military and AI startup Anthropic, following a social media announcement by Secretary Hegseth that prohibits any military-affiliated contractor, supplier, or partner from engaging in commercial activities with Anthropic. This decision comes after contentious negotiations over the use of Anthropic's AI models by the Pentagon, with the startup opposing their application for mass domestic surveillance or autonomous weapons. The Pentagon's demand for unrestricted use of the AI technology led to the designation of Anthropic as a supply chain risk, a move that allows the military to exclude certain vendors from defense contracts due to security concerns.

Anthropic has responded by threatening legal action, arguing that the designation sets a dangerous precedent for American companies negotiating with the government. The company claims it has not received direct communication from the Department of Defense or the White House regarding the negotiations. Anthropic also questions the legal authority behind Hegseth's statement, suggesting it lacks statutory backing.

The Pentagon has declined to comment on the situation, but the announcement has sparked widespread concern in Silicon Valley. Industry figures, including Dean Ball of the Foundation for American Innovation and Paul Graham of Y Combinator, have expressed alarm over the government's actions, viewing them as impulsive and potentially harmful to American innovation. OpenAI's CEO, Sam Altman, revealed that his company has reached an agreement with the Department of Defense that respects certain safety principles, contrasting with Anthropic's current predicament.

Anthropic's blog post clarifies that the supply chain risk designation, under 10 USC 3252, pertains only to direct Department of Defense contracts and does not affect how contractors use its AI software for other clients. Legal experts are uncertain about the immediate implications for Anthropic's customers, as the announcement lacks a clear legal foundation. Major tech companies like Amazon, Microsoft, Google, and Nvidia, which collaborate with both the military and Anthropic, have not commented on the matter.

Typically, supply chain risk designations require a thorough risk assessment and congressional notification before taking effect, but the current situation could deter other tech firms from engaging with the Pentagon. Greg Allen from the Center for Strategic and International Studies suggests that the Defense Department's actions may discourage companies from entering defense contracts due to potential overreach.

Legal experts anticipate that Anthropic will likely pursue litigation against the government, although such proceedings could take years to resolve, potentially impacting Anthropic's business in the interim. The dispute raises significant concerns for other US military partners, including Nvidia, Amazon, Google, and Palantir, which maintain close ties with Anthropic.

A tech executive, speaking anonymously due to the sensitivity of the situation, noted that their company is currently in a holding pattern, awaiting further directives from the Department of Defense. They compared the situation to Section 889 of the National Defense Authorization Act, which restricts federal contracts with companies using certain Chinese telecom equipment. The executive suggested that if the new mandate resembles this precedent, it could pose a substantial challenge for tech companies using Anthropic's AI technology.