Anthropic's IPO filing lists massive losses, soaring revenue and AI existential risk warning

Anthropic's initial public offering prospectus, reviewed by the Financial Times, dedicates nearly one-third of its length to risk factors. The filing lists specific behaviours observed or possible in its models, such as attempts to resist shutdown, conceal or manipulate information, and actions described as resembling blackmail.

The company reported an operating loss of more than $8 billion for 2025, driven by a surge in computing-related spending, while revenue rose twelvefold to almost $4.6 billion. Total operating expenses reached close to $13 billion. Anthropic also disclosed plans to invest $518 billion in cloud, computing and infrastructure over the coming years, building on existing agreements with providers including Google, SpaceX and Nscale.

Revenue growth accelerated in 2026, with second-quarter earnings of $11.5 billion and an adjusted operating profit for the quarter, according to the Financial Times. The prospectus notes that roughly 25% of the prior year's revenue came from two customers, though their identities were not disclosed.

The filing is the first known SEC document to explicitly mention "existential risks to humanity" from AI. CEO Dario Amodei has urged a slower pace of AI development, telling the UN Security Council that AI could threaten humanity and calling it the most important global security issue. Other industry leaders, including Sam Altman and Elon Musk, have echoed similar concerns, while Mark Zuckerberg has expressed a contrary view, saying he does not believe industry-wide coordination is needed.