Anthropic CEO resists Pentagon's push for full AI control
What's the story
Anthropic's CEO Dario Amodei has refused the Pentagon's request for unrestricted access to its artificial intelligence (AI) systems. The refusal comes as a response to Defense Secretary Pete Hegseth's ultimatum, which required compliance by Friday or face severe government action. Amodei stressed that "Anthropic understands that the Department of Defense, not private companies, makes military decisions."
AI concerns
Amodei highlights potential AI misuse scenarios
Amodei also raised concerns about the potential misuse of AI in certain scenarios. He said, "In a narrow set of cases, we believe AI can undermine, rather than defend, democratic values." The two scenarios he highlighted were mass surveillance of Americans and fully autonomous weapons with no human involvement. The Pentagon wants to use Anthropic's model for all lawful purposes without being dictated by a private company.
DOD tactics
Pentagon's attempts to force compliance
The Department of Defense has tried to force Anthropic into compliance by designating it a supply chain risk or invoking the Defense Production Act (DPA). The DPA allows the president to compel companies to prioritize national defense production. Responding to these threats, Amodei pointed out their contradiction: "One labels us a security risk; the other labels Claude as essential to national security."
Future plans
Amodei urges Pentagon to reconsider
Amodei said, "It's the Department's right to choose contractors most aligned with its vision, but given the substantial value that Anthropic's technology provides to our armed forces, we hope they reconsider." He also expressed Anthropic's willingness to continue serving the Department with certain safeguards in place. If offboarding is chosen by the Department, Amodei promised a smooth transition to another provider without disrupting ongoing military planning or operations.
AI contracts
Pentagon explores alternatives amid standoff with Anthropic
The Pentagon is currently negotiating contracts with several large language model providers, including Google, xAI, and OpenAI. These efforts are aimed at shaping future military applications of artificial intelligence in areas like autonomous drone swarms and cyber operations. Until recently, Anthropic was the only LLM developer on classified networks. However, this week the Pentagon announced a deal allowing xAI to operate in those environments.
Dispute escalation
Clarification on Maduro raid concerns
The dispute between Anthropic and the Pentagon escalated when officials raised concerns that Anthropic had questioned whether its products were used in a Venezuelan military raid that captured President Nicolas Maduro. However, Amodei clarified during the meeting that Anthropic had not raised concerns to Palantir or the Pentagon about the use of its AI tools in this operation. He also argued that existing safeguards would not hinder current Defense Department activities.