A heated showdown is unfolding between the Trump administration and artificial intelligence company Anthropic — and now Sen. Elizabeth Warren is stepping into the fight.

The Massachusetts Democrat is accusing President Trump and Defense Secretary Pete Hegseth of attempting to “extort” the AI company into loosening restrictions on how its technology can be used by the military. At the center of the dispute are guardrails Anthropic built into its AI system, Claude, that limit its use in fully autonomous weapons and large-scale domestic surveillance.

Warren, the top Democrat on the Senate Banking Committee, said the administration’s actions would strip away “common sense guardrails” that protect Americans from dangerous applications of AI, including weapons systems capable of operating without human oversight.

She called on Hegseth to testify before Congress and explain what she described as an “extraordinary abuse of power,” warning that the standoff could have serious national security consequences.

The conflict escalated after the Pentagon reportedly labeled Anthropic a “supply chain risk,” effectively barring U.S. military contractors and partners from working with the company. President Trump also directed federal agencies to immediately stop using Anthropic’s technology, publicly criticizing the firm and accusing it of attempting to force the Defense Department to follow its internal policies.

The dispute threatens a $200 million Pentagon contract awarded to Anthropic last summer. That agreement has now been thrown into uncertainty because of disagreements over how the AI model may be deployed.

Anthropic’s policy prevents its technology from being used for fully autonomous lethal weapons or mass surveillance of U.S. citizens. According to reports, the Pentagon has sought broader access to the AI system for “all lawful purposes,” a request the company’s leadership says it cannot accept.

Anthropic CEO Dario Amodei said the company made “virtually no progress” in negotiations and stated that agreeing to the government’s proposed terms would go against its principles.

The administration’s decision to classify Anthropic as a supply chain risk has drawn sharp criticism from other lawmakers. Sen. Ed Markey, also of Massachusetts, called the move “reckless and unprecedented,” arguing that such labels have historically been reserved for foreign adversaries — not American companies.

Anthropic, for its part, described the designation as “legally unsound” and warned that it sets a troubling precedent for how the government interacts with private technology firms.

The standoff highlights a broader debate playing out across Washington: how to balance national security interests with ethical safeguards in rapidly advancing AI systems. As artificial intelligence becomes increasingly embedded in defense and intelligence operations, questions about oversight, accountability and limits are intensifying.

With billions of dollars in defense contracts and the future of military AI development on the line, this battle is likely far from over. For now, it has sparked a larger conversation about who gets to set the rules for some of the most powerful technology ever created — the government, or the companies building it.