Anthropic is heading toward what could be a roughly $2 trillion IPO, and its paperwork is doing something unusual: it is openly worrying about the government.
The AI company has reportedly warned that shifting government policies toward artificial intelligence could disrupt its customer relationships and damage its reputation as it prepares for that potential listing, according to Reuters on Friday.
In its IPO prospectus, Anthropic said government actions and perceptions could lead to lost revenue, operational interruptions and harm involving customers, partners, employees and investors.
That is not a hypothetical worry. The company cited several encounters with the U.S. government over the past year. In February, an order directed federal agencies to stop using Anthropic's AI models, while the Department of Defense later designated the company a national security supply-chain risk.
Then, in June, the Department of Commerce imposed worldwide export restrictions on two Anthropic models. Commerce subsequently removed those restrictions, and Anthropic restored access, but the company cautioned that comparable actions could happen again.
Here is the part that keeps the risk in perspective: Anthropic said government contracts account for less than 1% of its annual revenue. So the direct dollars at stake are small. The reputational and relationship damage is the bigger question.
The prospectus also acknowledged that advanced AI systems could pose "catastrophic or existential risks to humanity." Which is a striking thing to put in a document designed to sell shares.
Anthropic and the White House did not immediately respond to MarketDash's request for comments.
