Pentagon vs. Anthropic: who will set the rules for military AI?

Pentagon vs Anthropic: Who Will Set the Rules for Military AI?

A tense conflict between the U.S. Department of Defense and startup Anthropic has reached a critical point, raising a sharp but crucial question: who should establish the framework for the use of artificial intelligence for military purposes—the executive branch, private companies, or Congress and a broader democratic process? The situation surrounding Anthropic's refusal to provide the Pentagon with unrestricted access to its AI systems, including the ability to use them for domestic surveillance and autonomous weapon guidance, has revealed deep disagreements over approaches to regulating this promising technology.

The conflict erupted when Defense Secretary Pete Hegseth reportedly gave Anthropic CEO Dario Amodei an ultimatum: either the company removes restrictions on the Department of Defense's use of its AI systems, or it faces consequences. Anthropic refused to cross two principled lines: allowing its models to be used for domestic surveillance of U.S. citizens and for fully autonomous military guidance. Hegseth, in turn, expressed disagreement with what he called "ideological restrictions" built into commercial AI systems, arguing that determining legitimate military applications should be the responsibility of government, not the supplier. "We will not use AI models that won't let you wage wars," he stated at a recent conference.

At its core, this dispute can be viewed as a disagreement over procurement. In a market economy, American military leadership determines what products and services it wants to purchase, and companies decide what they are willing to sell and on what terms. Neither side can be said to be wrong when it takes its position.

If a product does not meet operational needs, the government can turn to another supplier. If a company believes certain uses of its technology are unsafe, premature, or incompatible with its values or acceptable risk level, it can decline to provide them. For example, a coalition of companies signed an open letter committing not to use general-purpose robots for military purposes.

Such basic symmetry is an inherent feature of a free market.

The situation becomes more complex and troubling when it comes to designating Anthropic a "supply chain risk." This tool is intended to address genuine national security threats, such as actions by foreign adversaries. It should not be used to blacklist an American company for refusing to accept preferred contract terms offered by the government.

Using this power in such a manner represents a significant shift—from a procurement disagreement to the application of coercive leverage. Hegseth announced that "immediately, no contractor, supplier, or partner conducting business with the U.S.

Department of Defense can conduct any commercial activity with Anthropic." These actions will almost certainly face legal challenges, but they raise the stakes far beyond the loss of a single Pentagon contract.

It is also important to distinguish between two significant issues raised by Anthropic. The first—opposing domestic surveillance of U.S. citizens—touches on well-established rights and freedoms. The second—refusing autonomous weapon guidance—involves more complex ethical and practical considerations regarding accountability for military action and the risk of unintended escalation. Both of these issues require deep public discussion and legislative regulation, not back-room decisions between defense officials and technology companies.

This conflict underscores the acute need to establish a clear legislative framework and democratic oversight of the development and application of military artificial intelligence. Rather than relying on ultimatums and hidden mechanisms like designating a "supply chain risk," Congress should take the lead in developing transparent rules. These rules should ensure that the advancement of cutting-edge technologies in the military sphere does not occur at the expense of civil liberties and ethical principles. The balance between the necessity of maintaining defense capability and protecting fundamental citizen rights is critically important, and achieving it requires open dialogue and responsible legislative oversight, not unilateral decisions by the executive branch or technology giants.