Artificial intelligence company Anthropic is standing firm in its escalating dispute with the United States Department of Defense, declining to dilute internal safeguards that restrict military use of its technology, according to a source familiar with the discussions.
The standoff follows a high level meeting between Anthropic chief executive Dario Amodei and US Defence Secretary Pete Hegseth at the The Pentagon in Washington. The talks were convened to address a months long disagreement over whether the government may deploy Anthropic’s artificial intelligence systems without the company’s self imposed constraints.
Anthropic has refused to modify policies that prevent its systems from being used for autonomous weapons targeting or for domestic surveillance activities within the United States. Pentagon officials have reportedly argued that compliance with existing United States law should suffice, contending that additional corporate restrictions interfere with national security procurement.
According to the source, Defence Secretary Hegseth presented Anthropic with a stark ultimatum. The company could be designated a supply chain risk unless it revises its use policies, failing which the government may invoke statutory authority to compel changes. Anthropic has been given until Friday to respond.
The dispute raises profound legal questions at the intersection of defence procurement law, corporate governance and artificial intelligence ethics. At issue is whether a private technology provider can impose binding limitations on sovereign military use, or whether federal authority ultimately prevails where national security is invoked.