Anthropic’s IPO: Human extinction, AI hype and who pays the bill

Call it the great public reckoning: AI’s ultimate litmus test.

The techies’ pitch to retail investors is familiar: here’s a once-in-a-generation opportunity to get in early on the most transformative technology of the century, courtesy of the company behind Claude. Never heard of it? Oh, I guess you’re not an insider. But believe me when I say artificial intelligence will transform everything. As the argument goes, anyone who fails to invest now risks missing out on extraordinary returns.

Did I mention Anthropic is reportedly seeking a $1.5 trillion to $2 trillion valuation, even though their reported valuation was $965 billion in May? That’s up to $2,000,000,000,000. Most of us would be lucky to have five or six of those zeros in our bank accounts.

Sounds mighty fine until you start digging into the numbers and find that some of those grand claims continues to be based on the evidence of trust me, bro. The technology is revolutionary, the valuations are climbing, and prominent investors insist the future will be bigger than anyone can imagine. But buying into an initial public offering (IPO) requires more than promises about what might happen. Investors need to know whether the business can generate enough cash to justify the staggering sums being poured into it.

Anthropic’s preliminary IPO filing, submitted to the US Securities and Exchange Commission on Oct. 26, makes for impressive reading on revenue growth. The infrastructure commitments, however, are even yuuuuuger (bigger, for those not versed in Trumpspeak). The prospectus raises some serious questions: can the AI investment boom sustain itself without even more money? Or will retail investors ultimately be left shouldering the eye-watering cost of everything needed to power it?

Also, what does this mean for you, my fellow retail investor in Spain? Let’s dig in.

The AI payoff is still an open question

Anthropic’s revenue is growing. So are its obligations.

Anthropic’s S-1 filing shows revenue surging from $386 million in 2024 to $4.585 billion in 2025. In the first half of 2026, it reached $16.286 billion, up from $1.163 billion a year earlier.

Customers want its products, but building and running them comes at a steep price. For instance, compute and infrastructure operating expenses totalled $7.326 billion in 2025, exceeding annual revenue. In the first half of 2026, they reached $8.945 billion against $16.286 billion in revenue. The ratio is improving, but that’s not the same as proving the business is sustainable.

Then come the obligations. As of June 30, 2026, Anthropic had disclosed $265.6 billion in fixed or minimum guaranteed compute and infrastructure commitments, with contract terms of one to 20 years. It also reported $321.4 billion in estimated future lease obligations for signed leases that had yet to commence, primarily for computing infrastructure. These are separate categories and shouldn’t be treated as bills due today or simply added together.

The scale is so staggering, that minimum commitments can persist even if demand falls short. Anthropic had $112.4 billion in cash and cash equivalents and another $19.3 billion in short-term investments as of June 30, but if you’re seriously considering its IPO, you’ll be betting that customers keep spending and revenue grows fast enough to cover those infrastructure costs when the bills come due.

The circular financing problem

It may sound like a Ponzi scheme, but the circular relationships between chipmakers, AI developers, cloud providers and hyperscalers are well documented. Major technology companies can be investors in AI developers, suppliers of their computing capacity, distributors of their products and competitors building their own models.

Money moves through the ecosystem in several directions. An investor provides capital to an AI company, which spends some of it on cloud services. The cloud provider records that spending as revenue and can use its financial strength to invest further in AI. Those investments help finance more infrastructure spending, reinforcing expectations of future growth.

These are real transactions, and circular financing does not, by itself, imply fraud. But it can make the ecosystem more vulnerable if reported growth depends too heavily on interconnected investments and spending among a relatively small group of companies, rather than independent customer demand. The question is whether these businesses can generate attractive returns from customers willing to pay for AI products and services, without relying on the same network of investors and suppliers to keep the money flowing.

The Bank for International Settlements raised similar concerns in its January 2026 report, “Financing the AI boom: from cash flows to debt.” It warned that the scale of expected AI investment would require a shift from operating cash flows toward debt, with private credit playing a growing role. The sustainability of the AI boom, it argued, depends on companies meeting high earnings expectations.

If funding becomes more expensive or investors lose confidence, AI companies could cut spending, reducing revenue for infrastructure providers and weakening the value of strategic investments. A financing slowdown could ripple through the industry even if businesses continue to use AI.

Anthropic’s AI risks go beyond the IPO

Anthropic’s prospectus warns that sufficiently advanced AI systems could pose “catastrophic or existential risks to society” if not developed and deployed responsibly, including the possibility of human extinction (!!!). It’s a startling admission from a company seeking public investment to keep scaling the technology. The roughly 93 pages devoted to risk factors also cover competition, regulation, cybersecurity, legal liability and other potential harms that could undermine customer trust or future growth.

The risks extend beyond investors. Data centers require electricity, water, land and grid capacity. In Spain, where water scarcity and energy costs are already pressing concerns, AI infrastructure expansion has implications for households and businesses as well as shareholders. Generation Wealth has examined these pressures in its reporting on data centers, water shortages and energy bills in Andalusia.

AI could deliver productivity gains, create businesses and support better-paid jobs. But the benefits may be unevenly distributed, while workers face uncertainty over automation, wages and retraining costs. We have written about how AI growth impacts on employees and personal finances and what that could mean for people’s livelihoods.

Should Spanish investors buy Anthropic’s IPO?

It’s not every decade that a tech juggernaut goes public promising riches and abundance to anyone who gets on board. In Spain, where blockbuster IPOs are even rarer, the temptation to get in early can be all the stronger. But caution is warranted. Retail investors should understand how an IPO works, what the offer price implies and how easily a promising company can become an overpriced stock. We’ve warned about the risks of buying into IPO hype before, including in our guide at investing in IPOs such as SpaceX.

Anthropic could become one of the defining companies of the AI era, with genuine opportunities for growth. But buying its stock also means betting that AI adoption will keep expanding at a pace that justifies the price investors are asked to pay. A super intelligent technology and a fast-growing business don’t automatically make a compelling investment.

Perhaps the risk of human extinction is a small price to pay for the most revolutionary technology humanity has ever seen, one that promises to deliver abundance at last. To that end, we might not live to enjoy the fruits. Now, that’s a morbid thought.