Anthropic’s potential IPO could force investors to confront a question that goes far beyond one company: Are we in a classic bubble, collectively overvaluing the AI sector before we fully understand its risks, economics and long-term sustainability?
According to reporting from the Financial Times and Reuters, Anthropic devoted an unusually large portion of its IPO prospectus to risk factors, including disclosures that its models have demonstrated – or could demonstrate – behaviours involving attempts to “resist shutdown,” “conceal or manipulate information,” and behaviour resembling “blackmail.”
Those are extraordinary disclosures for a company potentially targeting a valuation above $2 trillion.
But the financial picture raises equally significant questions.
Explosive growth, enormous losses
Reuters reports that Anthropic generated approximately $4.6 billion in revenue in 2025, up roughly twelvefold, but recorded an operating loss exceeding $8 billion, with total operating expenses approaching $13 billion as computing costs surged.
And then comes the staggering capital requirement.
The prospectus reportedly outlines approximately $518 billion in spending on cloud, computing and infrastructure in the coming years.
Yet the growth story is undeniably remarkable. The Financial Times reports that Anthropic’s Q2 2026 revenue reached approximately $11.5 billion, putting the company on track for a second consecutive quarter of adjusted operating profit.
Extraordinary proposition
The investment proposition therefore looks extraordinary:
Explosive revenue growth
Potential multi-trillion-dollar valuation
Enormous infrastructure requirements
Billions in historical operating losses
Significant customer concentration, with nearly a quarter of last year’s revenue reportedly coming from just two clients
And then there is something few previous technology companies have had to discuss in their investor disclosures:
Potential risks to humanity
The $500 billion question
The question isn’t whether AI is transformative. It clearly is.
The question is whether current AI valuations are pricing in the upside while sufficiently accounting for the enormous costs, risks and uncertainties on the other side of the equation.
AI could reshape software, healthcare, finance, scientific research and virtually every knowledge industry. If frontier models become a general-purpose layer of the global economy, today’s leading AI companies could eventually justify extraordinary valuations.
But getting there requires extraordinary investment.
Compute needs to be built. Data centres need to be financed. Energy infrastructure needs to expand. Semiconductor supply chains need to scale.
If much of Anthropic’s reported $518 billion commitment is tied to long-term cloud and computing agreements that cannot easily be cancelled, investors are effectively underwriting a massive bet on future AI demand.
The assumption is that revenue will eventually catch up with infrastructure.
What happens if it doesn’t?
A referendum on AI
This is why Anthropic’s IPO could become much bigger than Anthropic.
If one of the world’s most prominent AI laboratories goes public while simultaneously warning investors about potentially severe model behaviours, extraordinary infrastructure requirements, customer concentration and enormous historical losses, public markets will have to decide how much uncertainty they are willing to pay for.
A successful IPO at an enormous valuation would validate the current AI investment thesis.
A disappointing reception could force investors to reconsider how much future growth is already priced into the sector.
Anthropic’s IPO may therefore not simply answer the question of what Anthropic is worth.
It could help answer a much larger question: What is the AI ecosystem really worth once we account for the capital required to build it, the risks involved in operating it, and the uncertainty surrounding its ultimate economic value?
That is why this IPO matters.
Anthropic isn’t just asking investors to buy into an AI company. It may be asking them to place a price on the future of AI itself.