Anthropic will not lift the safeguards that prevent its technology from being used for military targeting or domestic surveillance, according to Reuters, which cited a person familiar with the matter. The decision follows a meeting between Anthropic’s leadership and senior officials from the U.S. Department of Defense.
The company’s CEO, Dario Amodei, met directly with U.S. Defense Secretary Pete Hegseth to address a dispute that has been ongoing for months. At the core of the disagreement are Anthropic’s internal restrictions designed to block uses that would enable autonomous weapon targeting or U.S. domestic surveillance.
Representatives from the Pentagon have maintained that the government should be bound only by existing U.S. law. During the meeting, according to the person familiar with the discussions, Secretary Hegseth presented Anthropic with an ultimatum: be designated a supply-chain risk, or the government would resort to invoking a law that would compel the company to amend its usage rules. The government provided Anthropic with a response deadline of Friday.
The meeting was framed as an effort to resolve an extended disagreement between the tech firm and the Defense Department. Anthropic’s position, as reported, remains to uphold safeguards that prevent certain military and domestic surveillance applications of its AI systems.
The specifics of how the government would implement either course of action - designating Anthropic as a supply-chain risk or enforcing rule changes through legal authority - were not detailed in the account available from the person cited. The timeline established by the Pentagon for Anthropic’s reply is limited, with a response expected by the stated Friday deadline.
Given the information available, the interaction between the company and the Defense Department reflects continuing tensions over where private-sector safety measures intersect with government needs and legal frameworks. The account reports the positions and the ultimatum as described by a person familiar with the matter and relayed through Reuters.
Clear summary - Anthropic has refused to remove rules that bar its AI from being used for autonomous weapons targeting and U.S. domestic surveillance after a meeting between its CEO and the U.S. Defense Secretary. The Pentagon said the government should only have to follow U.S. law and gave Anthropic until Friday to respond to an ultimatum: accept being labeled a supply-chain risk or face legal compulsion to change its safeguards.