Anthropic Rejects $200M Pentagon Deal
AI safety startup Anthropic walked away from a $200M Department of Defense contract, refusing to accept restrictions related to mass surveillance and autonomous weapons. The move reportedly prompted the Pentagon to threaten classifying Anthropic as a "supply chain threat," showcasing the growing tension between AI ethics and lucrative government work.
The core of the dispute was the Pentagon's insistence on using Anthropic's AI for "all lawful purposes," a demand CEO Dario Amodei argued was overly broad. Anthropic specifically sought to prohibit the use of its technology for the mass surveillance of American citizens' data—including web browsing history and credit card records—and in fully autonomous weapons systems. The "supply chain risk" designation threatened by Defense Secretary Pete Hegseth is a classification historically reserved for foreign adversaries, not American companies. This move would not only bar the Department of Defense from doing business with Anthropic but could also prohibit any military contractor, supplier, or partner from engaging in commercial activity with the AI firm. Anthropic has stated it will challenge this unprecedented designation in court. Within hours of Anthropic's final rejection, rival AI lab OpenAI accepted a similar deal with the Pentagon. While OpenAI stated its agreement includes "guardrails" against mass domestic surveillance and the direction of autonomous weapons, Anthropic had previously rejected a similar architectural compromise, arguing that the modern military's mesh networks effectively erase the line between cloud servers and battlefield hardware. Founded in 2021 by former OpenAI researchers, Anthropic has built its identity and product, Claude, around a commitment to AI safety. This principle is embedded in its "Constitutional AI" framework, which aims to align the technology with human values and prevent harmful applications. The company's stance in the Pentagon negotiations reflects these foundational concerns. Financially, Anthropic is not reliant on government contracts, having raised billions from a diverse group of high-profile investors. The company recently closed a $30 billion Series G funding round, achieving a post-money valuation of $380 billion. Major backers include GIC, Coatue, Sequoia Capital, and significant investments from tech giants like Google and Amazon. The Pentagon contract was part of the Chief Digital and Artificial Intelligence Office's (CDAO) initiative to accelerate the adoption of advanced AI capabilities. The program awarded contracts with a ceiling of $200 million each to four major AI companies: Anthropic, Google, OpenAI, and xAI, aiming to develop AI workflows for various national security missions.
