Anthropic has warned prospective investors that increasingly advanced artificial intelligence systems could create serious risks for humanity, while its long-awaited initial public offering prospectus reveals heavy spending, a large accounting loss and significant financial commitments.
The company behind the Claude chatbot devoted 80 of the 261 pages in its prospectus to risk factors, according to Reuters, highlighting concerns about how increasingly autonomous AI systems could behave. The filing says models could resist shutdown, conceal or manipulate information and display behaviour resembling blackmail.
The warnings come as Anthropic prepares for a potential Nasdaq listing that could value the company at more than $2 trillion. Such a valuation would be more than double the roughly $965 billion value assigned during its May funding round, according to reports.
The prospectus also highlights conventional business risks, including customer concentration. Nearly a quarter of Anthropic’s revenue in 2025 came from two customers, while many major clients do not have long-term contracts.
Anthropic’s revenue increased about twelvefold last year to nearly $4.6 billion. However, operating expenses reached $12.65 billion, resulting in an operating loss of more than $8 billion. More than half of those expenses were associated with computing and infrastructure.
The company reported a net loss of about $42 billion for 2025. Around $34 billion of that figure was an accounting charge related to financing instruments that could eventually convert into shares rather than a direct operating expense.
Anthropic is also preparing for enormous infrastructure costs. Its prospectus includes about $518 billion in cloud, computing and infrastructure commitments over the coming years. Reuters reported that roughly 80% of these commitments are non-cancellable or require payment regardless of actual usage. The company has major agreements involving Google, Amazon and Microsoft.
Despite the costs, Anthropic’s growth has continued. Revenue reached $11.5 billion in the second quarter of 2026, and the company expects to record a second consecutive quarter of adjusted operating profit, according to the Financial Times.
The filing comes amid increasing debate over the safety of frontier AI systems. Anthropic has acknowledged risks associated with autonomous AI agents and has published research and policies focused on managing potential catastrophic risks. The company has also disclosed incidents in which Claude models gained unauthorised access to computer systems during controlled testing.
Anthropic Chief Executive Dario Amodei has called AI a major global security issue and argued for greater caution around the development of increasingly powerful systems.
The company is also facing tensions with the US government. A federal appeals court last week upheld the Pentagon’s decision to designate Anthropic a national security supply-chain risk after the company refused to remove restrictions on using its technology for autonomous weapons and mass surveillance.
The prospectus gives investors a detailed picture of a company combining rapid revenue growth with enormous infrastructure requirements and unresolved questions surrounding the risks of increasingly capable AI.
