Defense Secretary Summons Anthropic’s Amodei Over Military Use of Claude

When Defense Secretary Pete Hegseth picks up the phone to summon a tech CEO to the Pentagon, the conversation rarely ends with pleasantries. On Tuesday morning, Anthropic CEO Dario Amodei will find himself in exactly that position—a high-stakes meeting that could determine whether his company’s AI models remain in the U.S. military’s toolkit or get cast out entirely.

“The meeting comes as the Pentagon threatens to declare Anthropic a ‘supply chain risk’—a label typically reserved for foreign adversaries.” — Axios

The Ultimatum at the Pentagon

The tension between Anthropic and the Department of Defense has been building for months. Last summer, Anthropic signed a $200 million contract with the DOD, putting its Claude AI models at the center of the military’s emerging AI strategy. But the relationship soured when Anthropic refused to allow its technology to be used for mass surveillance of Americans and the development of autonomous weapons systems that could fire without human involvement.

The breaking point came in January, when Claude was reportedly used during a special operations raid that resulted in the capture of Venezuelan president Nicolás Maduro. The episode brought the philosophical divide between Anthropic and the Pentagon into sharp focus—one that Hegseth now appears determined to resolve, one way or another.

A High-Stakes Gamble

The supply chain risk designation is not an empty threat. If applied to Anthropic, it would void the company’s $200 million contract and force other Pentagon partners to drop Claude entirely. The label is typically reserved for foreign adversaries like Huawei or Kaspersky—companies deemed too risky to touch. For an American AI firm, it would be unprecedented.

Replacing Anthropic would be a significant undertaking. The military has integrated Claude into various operations, and finding a replacement with similar capabilities would take time and resources. But Hegseth appears willing to make that call if Amodei doesn’t budge on the company’s ethical restrictions.

The broader implications extend far beyond this single contract. Other AI labs are watching closely. If the Pentagon successfully pressures Anthropic into relaxing its safety guidelines, it could set a precedent that ripples through the entire industry. Conversely, if Anthropic holds its ground and survives the confrontation, it could embolden other companies to push back against military demands.

“Hegseth is giving Amodei an ultimatum: play ball or be banished. It’s unclear whether he’s bluffing.” — Axios source

The AI Ethics Battleground

Anthropic has built its reputation on safety-first AI development. The company’s constitutional AI approach and public commitments to responsible deployment have made it a favorite among policymakers and researchers who worry about the risks of unchecked AI capabilities. But those same commitments have put it on a collision course with military users who want fewer restrictions, not more.

The specific points of contention—mass surveillance and autonomous weapons—touch on some of the most contentious issues in AI ethics. Surveillance raises Fourth Amendment concerns and questions about civil liberties. Autonomous weapons, meanwhile, have been the subject of international debate, with many experts calling for outright bans on systems that can kill without human decision-making.

For Amodei, the Tuesday meeting represents more than a business negotiation. It’s a test of whether an AI company can maintain its ethical commitments while working with the world’s most powerful military. The outcome could shape how AI firms engage with defense contracts for years to come.

What Comes Next

The immediate question is whether Hegseth follows through on his threat. Declaring Anthropic a supply chain risk would be an aggressive move with significant consequences for both sides. The military would lose access to a capable AI system, and Anthropic would lose a major revenue stream—not to mention the reputational damage of being labeled a security risk.

But the longer-term question is how AI companies navigate the growing demand from military and intelligence agencies. As AI capabilities advance, the temptation to deploy them in sensitive contexts will only increase. The companies that build these systems will face increasing pressure to compromise on safety guidelines in exchange for lucrative contracts.

For now, all eyes are on the Pentagon. When Amodei walks into that meeting on Tuesday morning, he’ll be carrying more than Anthropic’s future—he’ll be carrying a piece of the AI industry’s soul.

This article was reported by the ArtificialDaily editorial team. For more information, visit TechCrunch.