Plus, Nvidia wants to extend its AI chip reach Many tech companies survive without contracts with the federal government. None are planning to go public later this year at a valuation hopefully well north of $380 billion—while still burning loads of cash.
Anthropic’s standoff with the Pentagon has snowballed into the AI company being effectively banned from doing business with any federal agency, at least according to a missive fired off by President Trump on Friday. That followed an apparently tense face-to-face meeting between Anthropic CEO Dario Amodei and Defense Secretary Pete Hegseth, who later designated Anthropic as a supply-chain risk.
That designation could impair Anthropic’s ability to work with any company that does business with the federal government. Anthropic says it will challenge that designation in court.
But even if successful on that legal front, Anthropic will likely have to go forward without the U.S. government as a customer. Guggenheim analyst John DiFucci estimates the U.S. federal government accounts for about 8%-9% of worldwide spending on software, which makes it difficult for any software vendor to ignore it as a customer.But Anthropic isn’t your typical software vendor. The company’s Claude AI tools are widely used across large and small businesses for tasks like coding automation. That success has made it into one of the most important names in AI—Anthropic’s software was being used by the U.S. government as part of the Iran attacks just hours after the ban was announced, the WSJ reported.
The company’s announcements of new services and capabilities have been the major spark behind the “SaaSpocalypse.” That has been a brutal selloff of software stocks that has cost companies like Salesforce, Workday and ServiceNow more than one-quarter of their market capitalizations so far this year.
Anthropic, in other words, already has its hooks deep into corporate America. And the company’s resistance to the Pentagon’s demands also seems to have rallied strong support among consumers, with the Claude app hitting number one in Apple’s App Store. It also surpassed OpenAI’s ChatGPT in that ranking for the first time.
Tech companies have taken seemingly costly stands before and stuck by them. Google famously exited the Chinese internet search market in 2010 after deciding it wasn’t comfortable censoring results to abide by the government’s censorship rules. But Google was already public at that point, very profitable and generated about $8.5 billion in annual free cash flow.
Anthropic, by contrast, is reportedly burning through about $3 billion a year in cash. Losing federal-government contract dollars, and possibly more from government contractors, could dent the company’s efforts to cast itself as a stronger business than arch-rival OpenAI. The Wall Street Journal reported late last year that the company was aiming to turn a profit ahead of the ChatGPT maker, which doesn’t expect to get to that level until 2030.
Pulling that off without Uncle Sam’s support will be much tougher. It’s IPO might be as well.
