‘attempted Corporate Murder’: Trump’s Threats Against Anthropic Chill Ai Industry

The high-stakes standoff between the Trump administration and artificial intelligence startup Anthropic is sparking fears in Silicon Valley, on Capitol Hill and across K Street of a fundamental shift in the balance of power between Washington and the AI industry.

President Donald Trump’s fiery attack on the company escalated the feud on Friday, as he ordered a government-wide boycottof Anthropic’s Claude AI model and threatened the company with prosecution if it doesn’t cooperate with agencies winding down their use of the technology. Within minutes, Defense Secretary Pete Hegseth also designated Anthropic a supply chain risk — a label, typically reserved for foreign firms with ties to U.S. adversaries, that blocks companies with government contracts from working with Anthropic.

Trump was silent on another threat that Hegseth had delivered to the company earlier this week, when he suggested DOD would invoke the Cold War-era Defense Production Act to force the company to let the government use Claude.

The dispute has opened up a wider debate on AI’s potential for enabling mass surveillance of U.S. citizens and computer-directed killing, two behaviors that Anthropic CEO Dario Amodei said his company would not allow in its work with the Pentagon.

But the administration’s actions also threaten to cripple Anthropic, beyond just wiping out its $200 million Pentagon contract. And the threats leveled by the government against the company this week pose an ominous development for the entire AI industry, legal experts told POLITICO.

Any effort to invoke the Defense Production Act against Anthropic would cause “an effective partial nationalization of the AI industry,” said Alan Rozenshtein, a professor at the University of Minnesota Law School and former Justice Department official who specializes in technology law.

Rozenshtein said it was inevitable that Washington would eventually move to tighten its grip on the tech sector, given the rapid growth in the capabilities of the most powerful AI models.

“For years, people have said, ‘Look, it is just not realistic that we are going to have private companies develop the Machine God, and then the government is just going to sit there and say, ‘Oh, that’s interesting.’ We were always going to nationalize,’” said Rozenshtein. “And this is the way you do it. Under American law, DPA is probably the most straightforward way of doing it.”

Other AI companies are waking up to the threat.

In an internal company memo published on Thursday and obtained by POLITICO, OpenAI CEO Sam Altman — whose company also contracts with the Pentagon — said he is “especially concerned about the threat to invoke rarely used authorities like the DPA.” Altman said the dispute was “no longer just an issue between Anthropic and the [DOD]; this is an issue for the whole industry.”

Altman said OpenAI also opposes the Pentagon’s use of its ChatGPT model to surveil Americans or enable autonomous weapons systems. But he said it was important for all parties to deescalate, warning that the standoff “risks our national security, and also risks the government resorting to actions which could risk American leadership in AI.”

A spokesperson for OpenAI did not provide comment when asked if it has shared its red lines with the Pentagon, or if the company is facing the same pressure as Anthropic. Spokespeople for Google and xAI, both of which also contract with the Pentagon, did not respond to requests for comment.

Trump’s broadside Friday against Anthropic on Truth Social sent mixed signals on whether such a DPA-enforced government expropriation of a private company’s intellectual property might be in the offing.

“I am directing EVERY Federal Agency in the United States Government to IMMEDIATELY CEASE all use of Anthropic’s technology,” Trump wrote. “We don’t need it, we don’t want it, and will not do business with them again!”

But then he added that he was ordering agencies to begin a “Six Month phase out” of Anthropic’s products, adding: “Anthropic better get their act together, and be helpful during this phase out period, or I will use the Full Power of the Presidency to make them comply, with major civil and criminal consequences to follow.”

Trump didn’t specify what those “consequences” might be, and neither the Pentagon nor the White House immediately responded to requests to elaborate.

But Rozenshtein said he believes the Defense Production Act remains on the table, “especially with this administration that is not particularly rule-of-law constrained.” Violations of the Defense Production Act would result in criminal penalties, which Trump alluded to in his Friday post.

In his follow-up on X, Hegseth announced that the Defense Department was declaring Anthropic to be a “supply chain risk,” writing: “Effective immediately, no contractor, supplier, or partner that does business with the United States military may conduct any commercial activity with Anthropic.”

That means multitrillion-dollar tech giants such as Nvidia, Amazon and Google would need to cut ties with Anthropic, former Trump administration AI adviser Dean Ball wrote Friday night on X.

“This is simply attempted corporate murder,” Ball wrote. “I could not possibly recommend investing in American AI to any investor; I could not possibly recommend starting an AI company in the United States.”

On Friday, lawmakers on Capitol Hill sounded the alarm over the standoff.

In a bipartisan letter, four lawmakers in charge of defense policy — Senate Armed Services heads Roger Wicker (R-Miss.) and Jack Reed (D-R.I.), along with top Senate defense appropriators Mitch McConnell (R-Ky.) and Chris Coons (D-Del.) — urged Hegseth and Amodei to extend negotiations beyond the Pentagon’s “hasty” deadline of 5:01 p.m. Friday..

“Simply put, there is no reason to transform this situation into an all-or-nothing moment,” the senators wrote. If the dispute continued to escalate, they warned of harm to national security and the willingness of the tech industry to work with Washington.

Sen. Elizabeth Warren (D-Mass.) urged hearings on the AI powers at the heart of the dispute.

“Is the Trump administration punishing Anthropic because it’s refusing to help mass surveil American communities or build killer robots?” she wrote on X, calling for Hegseth to testify to Congress. “The American people deserve to know what Trump officials are planning at the Pentagon.”

Administration officials have denied that it intends to use Anthropic’s software for either purpose, but also insisted that the Pentagon could not tolerate any restrictions on how it deploys AI, short of having to follow U.S. law.

“Our position has never wavered and will never waver: the Department of War must have full, unrestricted access to Anthropic’s models for every LAWFUL purpose in defense of the Republic,” Hegseth said in his Friday post. He accused the company of “a cowardly act of corporate virtue-signaling that places Silicon Valley ideology above American lives.”

In a blog post Thursday rejecting the Pentagon’s demands, Amodei wrote that Anthropic “understands that the Department of War, not private companies, makes military decisions. … However, in a narrow set of cases, we believe AI can undermine, rather than defend, democratic values.”

He added that Anthropic would “enable a smooth transition to another provider” should the Defense Department choose to part ways.

Elsewhere in Washington, the risk-averse technology lobby also waded into the fight.

Chris Mohr, president of the Software and Information Industry Association, which counts Anthropic as a member, issued a statement Friday supporting the company’s request for narrow limits on the use of its model by the military. And he issued a warning against invoking the Defense Production Act — calling it an “overbroad response to a technical disagreement” — or blacklisting Anthropic, noting such an approach is “typically reserved for hostile foreign entities” and is “both untethered from the facts of Anthropic’s security posture and unlikely to advance a long-term solution.”

An Anthropic spokesperson did not immediately respond when asked if the company plans to sue over its designation as a supply chain risk. If it does, Rozenshtein expects a judge to rule in the company’s favor.

“The law pretty clearly does not apply to a circumstance like this,” said Rozenshtein, who noted the label has been almost exclusively reserved for foreign companies like Huawei, a Chinese telecom giant with close ties to Beijing.

“It clearly, clearly was not designed for an American company that has a contract dispute with the government,” said Rozenshtein.

But he added that it would be harder for Anthropic to defend against the government’s invocation of the Defense Production Act, given the broad national security powers it provides to the president. If the Trump administration took that shot, Rozenshtein said, Anthropic would have no choice but to fight back.

“This is an existential moment for this company,” said Rozenshtein. “They will use every tool in their toolbox.”