The U.S. Department of Defense has formally notified artificial intelligence firm Anthropic that it has been designated as a potential “supply chain risk,” a classification that could significantly restrict the company’s ability to conduct business with the U.S. government.

In a public statement released online on Thursday, Anthropic CEO Dario Amodei confirmed that the company had received an official notice from the Pentagon regarding the designation. He also reiterated the company’s position that it intends to challenge the decision through legal channels.

Amodei stated that Anthropic does not believe the government’s action is legally justified and indicated that the company will pursue all appropriate options to contest the designation.

Despite the growing dispute, Anthropic’s artificial intelligence technology is currently being used by the U.S. military during ongoing operations connected to a broader conflict involving Iran, according to two individuals familiar with the situation who spoke anonymously. The company’s systems analyze large volumes of data and satellite imagery collected by the United States. These insights help military planners determine troop deployments and evaluate potential strike locations.

The conflict between Anthropic and the Pentagon has intensified in recent weeks, particularly over the terms governing how the company’s AI tools could be deployed on classified defense systems. The Department of Defense insisted that Anthropic’s technology must be available for any lawful government use. Officials warned that failure to agree to these conditions would lead to the company being excluded from future government contracts.

Anthropic pushed back against those demands, arguing that its technology should only be used under conditions that prevent potential misuse. The company requested contractual guarantees that its AI would not be applied to domestic surveillance activities targeting American citizens or integrated into autonomous lethal weapons systems.

Defense officials rejected those restrictions, maintaining that private technology companies cannot impose limits on how the government employs tools in matters related to national security. From the Pentagon’s perspective, such decisions must remain under the authority of the U.S. government rather than individual contractors.

Negotiations between the two sides continued until a Friday afternoon deadline set by the Department of Defense. When the deadline passed without an agreement, Defense Secretary Pete Hegseth announced on social media that Anthropic would be classified as a supply chain risk to national security.

This classification is typically reserved for companies believed to have connections to the Chinese government or entities viewed as posing strategic risks. In the same statement, Hegseth also warned that any contractors, suppliers, or partners working with the U.S. military would not be permitted to maintain commercial relationships with Anthropic.

Anthropic had previously indicated that it would only pursue legal action after receiving a formal written notice confirming the designation. Following receipt of that letter, the company signaled that a legal challenge is now under consideration.

In his public message, Amodei also said that Anthropic remains engaged in discussions with defense officials even as the Pentagon has announced plans to remove the company’s technology from its systems within six months.

He added that the company intends to assist with the transition process. According to Amodei, Anthropic is prepared to continue providing its AI models to the Department of Defense and the broader national security community at minimal cost, while also offering engineering support to ensure an orderly replacement of the technology. The company says it will maintain that support for as long as it is permitted to do so.

At present, Anthropic is the primary provider of artificial intelligence technology used in classified Pentagon systems. However, several competing companies are already positioning themselves to fill the potential gap if the company is removed from government infrastructure.

Among those competitors are OpenAI and xAI, the artificial intelligence company founded by Elon Musk. Both firms have recently entered into agreements with the Department of Defense to supply AI tools designed for use on classified government platforms.

OpenAI announced its own deal with the Pentagon on Friday, shortly after President Donald Trump issued an order directing federal agencies to discontinue the use of Anthropic technology within six months.

Unlike Anthropic, OpenAI agreed to terms that allow the U.S. military to deploy its artificial intelligence systems for any lawful purpose. At the same time, the company said it had negotiated safeguards intended to preserve its internal safety standards. These safeguards involve technical controls designed to prevent certain high-risk uses of its AI systems.

The agreement quickly attracted criticism from some observers and policy analysts over the weekend, particularly concerning concerns about surveillance and civil liberties.

Responding to those concerns, OpenAI announced on Monday that it had revised its agreement with the Pentagon to include additional protections. According to the company, these new measures are intended to reduce the risk that its technology could be used for large-scale surveillance of American citizens.

Even with those changes, some critics argue that the revised agreement still leaves room for interpretation and may allow certain surveillance-related uses under specific conditions. The debate highlights the growing tension between national security priorities and the ethical limits technology companies seek to place on powerful artificial intelligence systems.

As artificial intelligence continues to play a larger role in military operations, disputes like the one between Anthropic and the Pentagon illustrate the complex challenges governments and private technology firms face when balancing innovation, security, and civil liberties.