The past two weeks have seen a clash between Anthropic CEO Dario Amodei and Defense Secretary Pete Hegseth as the two battle over the military’s use of AI.

Anthropic refuses to allow its AI models to be used for mass surveillance of Americans or for fully autonomous weapons that carry out strikes without human intervention. At the same time, Secretary Hegseth argued that the Department of Defense should not be limited by a vendor’s rules, arguing that any “lawful use” of the technology should be permitted.

On Thursday, Amodei publicly signaled that Anthropic was not backing down – despite threats that his company could be designated a supply chain risk as a result. But with the rapidly changing news cycle, it’s worth revisiting exactly what’s at stake in the fight.

At the heart of this fight is who controls powerful AI systems: the companies that build them or the government that wants to deploy them.

What is Anthropic worried about?

As we said above, Anthropic does not want its AI models to be used for mass surveillance of Americans or for autonomous weapons without any humans involved in targeting and shooting decisions. Traditional defense contractors typically don’t have a say in how their products will be used, but Anthropic has argued since its inception that AI technology presents unique risks and therefore requires unique safeguards. From the company’s perspective, the question is how to maintain these safeguards when the technology is used by the military.

The U.S. military already relies on highly automated systems, some of which are deadly. The decision to use lethal force has always been left to humans, but there are few legal restrictions on the military use of autonomous weapons. The DoD does not outright ban fully autonomous weapon systems. According to a 2023 DOD directive, AI systems can select and engage targets without human intervention, provided they meet certain standards and are reviewed by senior defense officials.

This is precisely what makes Anthropic nervous. Military technology is secret by nature, so if the U.S. military took steps to automate lethal decision-making, we might not know until it was operational. And if he used Anthropic’s models, that could be considered “lawful use.”

Anthropic’s position is not that such uses should be permanently excluded. This is because its models are not yet sufficiently efficient to support them safely. Imagine an autonomous system misidentifying a target, escalating a conflict without human authorization, or making a deadly split-second decision that no one can reverse. Put a less capable AI in charge of weapons and you get a very fast, very confident machine that is bad at making high-stakes calls.

AI also has the power to increase legal surveillance of American citizens to a worrying degree. Under current US laws, surveillance of US citizens is already possible, whether through the collection of text messages, emails and other communications. AI changes the equation by enabling automated large-scale pattern detection, entity resolution across datasets, predictive risk scoring, and continuous behavioral analysis.

What does the Pentagon want?

The Pentagon’s argument is that it should be able to deploy Anthropic’s technology for any legal use it deems necessary, rather than being limited by Anthropic’s internal policies on things like autonomous weapons or surveillance.

Specifically, Secretary Hegseth argued that the Department of Defense should not be constrained by a vendor’s rules and would engage in “lawful use” of the technology.

Sean Parnell, the Pentagon’s chief spokesman, said in an article published Thursday that the department has no interest in conducting mass domestic surveillance or deploying autonomous weapons.

“Here’s what we’re asking: Allow the Pentagon to use Anthropic’s model for any legal purpose,” Parnell said. “This is a simple, common-sense request that will prevent Anthropic from jeopardizing critical military operations and potentially putting our warfighters in harm’s way. We will not let ANY company dictate the terms of how we make operational decisions.”

He added that Anthropic has until 5:01 p.m. ET Friday to decide. “Otherwise, we will end our partnership with Anthropic and consider them a risk to DOW’s supply chain,” he said.

Despite DoD’s position that it simply does not believe it should be constrained by a company’s use policies, Secretary Hegseth’s concerns about Anthropic have at times seemed linked to cultural grievances. In a speech to the offices of SpaceX and xAI in January, Hegseth railed against “woke AI” in a speech that some saw as a preview of his feud with Anthropic.

“The War Department’s AI will not be awakened,” Hegseth said. “We build weapons and systems ready for war, not chatbots for an Ivy League college lounge.”

And now ?

The Pentagon has threatened to either declare Anthropic a “supply chain risk” – which prevents Anthropic from doing business with the government – ​​or invoke the Defense Production Act (DPA) to force the company to adapt its model to military needs. Hegseth gave Anthropic until 5:01 p.m. Friday to respond. But as the deadline approaches, no one knows whether the Pentagon will carry out its threat.

This is not a fight from which either side can easily walk away. Sachin Seth, a VC at Trousdale Ventures who focuses on defense technology, says a supply chain risk label for Anthropic could mean “extinction” for the company.

However, he added, if Anthropic is removed from DoD, it could pose a national security issue.

“[The Department] “It would take six to 12 months for OpenAI or xAI to catch up,” Seth told TechCrunch. “That leaves a window of up to a year in which they could work not on the best model, but on the second or third best.”

xAI is preparing to go classified and replace Anthropic, and it’s fair to say, given owner Elon Musk’s rhetoric on the matter, that the company would have no problem giving the DoD full control over its technology. Recent reports indicate that OpenAI may stick to the same red lines as Anthropic.