Anthropic managers say that current and potential customers have been demanding novel terms and even withdrawing from negotiations since the U.S. Department of Defense deemed the AI startup a supply chain risk overdue last month, according to court documents that also revealed novel financial details about the company.
Hundreds of millions of dollars in expected revenue from Pentagon-related work this year are already at risk for Anthropic, the company’s chief financial officer Krishna Rao wrote in a report. court application on Monday. But he said that if the government has its way and pressures a wide range of companies not to do business with the artificial intelligence startup, regardless of any ties to the military, Anthropic could ultimately lose billions of dollars in sales, he said. According to Rao, its total sales since commercializing the technology in 2023 have exceeded $5 billion.
Anthropic’s revenues skyrocketed as Claude models began to outperform competitors and demonstrate advanced capabilities in areas such as software code generation. However, the company spends heavily on computer infrastructure and remains deeply unprofitable. Rao specified that Anthropic has spent over $10 billion training and implementing its models.
Anthropic’s chief commercial officer, Paul Smith, gave several examples of partners who have privately raised concerns with the AI startup in recent days. He said a financial services client had halted negotiations on a $15 million deal because of a supply chain etiquette, and two leading financial services companies had refused to enter into a deal worth a combined $80 million unless they were given the right to unilaterally cancel the contracts for any reason. A grocery chain canceled a sales meeting, citing a supply chain risk flag. Smith added.
“They have all taken steps that reflect a deep distrust and growing fear of contact with Anthropic,” Smith wrote.
The executives’ comments are part of statements by six Anthropic leaders in support of the preliminary order, which will allow the San Francisco-based company to continue working with the Department of Defense until lawsuits over supply chain risk issues are resolved.
Anthropic has sued the Trump administration in two courts. The lawsuit, filed Monday in federal court in San Francisco, accuses the government of violating the company’s right to free speech. A separate case filed Monday in a federal appeals court in Washington accuses the Department of Defense of unfair discrimination and retaliation against Anthropic.
The company is seeking a hearing in San Francisco as early as Friday for a ephemeral waiver. The legal battle and degenerating sales follow a weeklong dispute between Anthropic and the Pentagon over the potential operate of artificial intelligence technology for mass domestic surveillance and autonomous lethal weapons. Anthropic argues that AI is not yet capable of safely performing tasks, while the Pentagon wants the right to make that assessment on its own.
By law, the supply chain designation prevents a petite group of companies that work with the Pentagon from including Anthropic in their systems. Defense Secretary Pete Hegseth, however, cast a wider net. He sent X announced overdue last month that “effective immediately, no contractor, supplier or partner doing business with the United States military may conduct any commercial activity with Anthropic.”
Rao wrote that the Pentagon reinforced that message by contacting several startups about their operate of Claude. He said it happened during a conversation with an investor shared by Anthropic and smaller companies. “They became worried and unsure about their ability to take advantage of Claude,” Rao wrote.
The Pentagon declined to comment on the lawsuits and did not immediately respond to a request for comment on Rao’s allegations about the coverage.