Anthropic’s IPO filing reads like a warning to its own buyers

Anthropic has done something unusual for a company chasing a market debut of historic size: it has told investors that the technology it sells could, in its own telling, threaten humanity itself. That warning appears in its long-awaited initial public offering prospectus, and it is not tucked away in a footnote for the bored and the brave.

Nearly a third of the S-1 is devoted to risk factors. Among them is the possibility that more advanced AI models could manipulate, blackmail or behave in other unpredictable ways, while Anthropic’s own research has shown autonomous models sabotaging code, assisting fraud and manipulating information in controlled tests. The company behind Claude is preparing for a Nasdaq listing expected this autumn, with backers hoping for a valuation above $2 trillion.

The financial disclosures are hardly more comforting. Close to a quarter of Anthropic’s revenue last year came from just two customers, and many of its largest clients are not tied to long-term contracts. Revenue jumped twelvefold to almost $4.6 billion in 2025, but operating expenses reached $12.65 billion, more than half of that on computing and infrastructure, leaving an operating loss of more than $8 billion.

The net loss came to about $42 billion, including a roughly $34 billion accounting charge linked to financing that could eventually convert into shares rather than cash spent on running the business. Anthropic says it plans to spend $518 billion on cloud, computing and infrastructure obligations in the coming years. At the end of 2025, it had $20.28 billion in cash and short-term investments. By the second quarter of 2026, revenue had reached $11.5 billion, and the company was on course for a second straight quarter of operating profit, albeit on an adjusted basis.

That is the oddity of this listing: a company valued like a future giant, while describing a balance sheet and technology stack that still look distinctly fragile. Investors are being asked to price in both the promise and the peril, which is a neat trick if you can pull it off. Anthropic’s backers expect a value above $2 trillion, more than double the $965 billion set in its funding round in May. At one point, pre-IPO contracts on the crypto platform Hyperliquid were already implying roughly $2.36 trillion.

The debate is not confined to spreadsheets. Current and former Anthropic staff have warned that runaway AI could end humanity within a decade, and chief executive Dario Amodei told the UN Security Council last week that AI was the most important global security issue facing the world today. He has also called this month for slowing the pace of frontier AI development, a view backed by OpenAI’s Sam Altman and Elon Musk.

Washington, predictably, has its own agenda. US President Donald Trump rejected Amodei’s call to slow down, saying he was for going ahead and winning, and describing the contest as a race between the US and China that America cannot afford to lose. The dispute between Anthropic and the administration has been building since February, when Anthropic refused to let its models be used for fully autonomous weapons or domestic mass surveillance. That led the Pentagon to classify the company as a supply chain risk, a decision a federal appeals court upheld last week.

So the prospectus lands exactly where one might expect it to: between a feverish market and a nervous policy debate, with neither side showing much appetite for restraint. Anthropic wants public investors to fund the next phase of AI expansion, while also warning that the technology may become difficult to control. The modern listing, it seems, now comes with a side order of existential risk.

Written by Freya Stensrud freya.stensrud@alpineweekly.com