•
•
•
•
•
•
•
Anthropic’s planned stock-market debut is drawing global attention after its IPO prospectus revealed both the extraordinary financial scale of the AI business and unusually stark warnings about the potential dangers of increasingly powerful AI systems.
The company behind Claude AI is seeking a public-market valuation that could exceed $2 trillion, according to Reuters. That would be more than double Anthropic’s roughly $965 billion valuation from its May funding round. However, the prospectus also highlights the enormous costs involved in building and operating frontier AI models.
Anthropic’s revenue grew roughly 12 times in 2025 to nearly $4.6 billion, demonstrating the rapid commercial adoption of its AI products. Yet the company reported a $42 billion net loss for the year.
A large part of that headline loss came from an approximately $34 billion accounting charge related to the changing estimated value of financing instruments that could eventually convert into Anthropic shares. Separately, its operating loss widened to about $8.06 billion, compared with $2.98 billion in 2024.
The company spent $7.33 billion on computing and infrastructure in 2025, around three times its spending a year earlier. Its prospectus also disclosed approximately $518 billion in future cloud, computing and infrastructure obligations, underlining the capital intensity of frontier AI development.
The financial disclosures are accompanied by an extraordinary section on AI safety and risk.
Anthropic's prospectus warns that increasingly advanced AI models could potentially cause “catastrophic or existential” harm. The filing discusses scenarios in which models could behave unexpectedly, including attempts to resist shutdown, conceal information, manipulate information or display behaviour resembling blackmail.
The prospectus reportedly devotes around 80 pages of its 261-page main body to risk factors, nearly twice the space devoted to describing the company's business. Anthropic also acknowledges that safety work is resource-intensive and competes with spending on computing capacity and AI talent.
These disclosures do not mean Anthropic is predicting that such outcomes will occur. Rather, they are risks the company is required to disclose to prospective investors as part of its IPO documentation.
Anthropic's IPO could become an important reference point for how public investors value leading AI companies. The company is competing with OpenAI, Google, Meta and xAI for customers, researchers and computing infrastructure. OpenAI is also preparing for a potential public listing.
The prospectus further highlights customer concentration as a business risk: nearly one-quarter of Anthropic's 2025 revenue came from two customers, while many major customers do not have long-term contracts that guarantee future spending.
Anthropic's public debut therefore arrives at a pivotal moment for the global AI industry. Its filing presents two sides of the AI boom: rapidly expanding demand and enormous valuations on one hand, and extraordinary computing costs, financial losses and unresolved AI-safety challenges on the other.
The IPO is expected to take place after the November US midterm elections, according to Reuters' reporting, although the final timing and valuation remain subject to the offering process.
