Anthropic’s IPO prospectus contains some of the starkest warnings ever filed by a company about to go public. Here is what the disclosures mean.

The headline warning: the company says its own AI could pose “catastrophic or existential risks to humanity”. This is an admission that the technology it profits from may be dangerous at scale.

The business warning: Anthropic says government attitudes toward it could hurt not just its state contracts, which are under 1 percent of revenue, but its commercial customers and partners too.

The examples: in February, US President Donald Trump ordered agencies to stop using its models; the Pentagon called it a supply-chain risk; in June, export curbs hit its Fable 5 and Mythos 5 models.

The money at stake: CFO Krishna Rao estimated hundreds of millions of 2026 revenue at risk, potentially multiple billions. Yet the company is still targeting a $2 trillion valuation.

The context: CEO Dario Amodei wants the industry to slow down, even as his company races to go public. Investors must decide whether the warnings are caution or a red flag.