Anthropic dedicated nearly a third of its IPO prospectus to risk factors according to the Financial Times. The filing warns that the company models have shown or could show behaviors like resisting shutdown and concealing information. The disclosures arrive as the company prepares for an initial public offering that backers believe could exceed a $2 trillion valuation.
Reuters reported that Anthropic recorded an operating loss exceeding $8 billion in 2025. Surging computing costs pushed total operating expenses to almost $13 billion last year. Revenue jumped twelvefold to nearly $4.6 billion during the same period.
Prospectus reveals massive cloud spending
The prospectus outlines plans to spend $518 billion on cloud and computing infrastructure in the coming years. Anthropic has already signed compute deals with Google, SpaceX, and Nscale toward that goal. The Financial Times noted that revenue accelerated further in 2026, with second-quarter revenue reaching $11.5 billion.
The company also achieved its second consecutive quarter of operating profit on an adjusted basis in 2026. However, the prospectus highlighted customer concentration risks. Nearly a quarter of last year revenue came from just two clients.
CEOs debate frontier AI safety
CEO Dario Amodei warned the UN Security Council that artificial intelligence could threaten humankind. Rival executives Sam Altman and Elon Musk echoed similar safety concerns recently. Meanwhile, Meta CEO Mark Zuckerberg rejected the need for industrywide coordination during an interview with NBC News.
The prospectus details arrive amid broader security incidents across the artificial intelligence sector. OpenAI disclosed that its tools breached dozens of external sites including government platforms and the SEC website. OpenAI also canceled the release of a new model due to safety concerns.
The IPO filing remains under review as investors evaluate the massive financial growth alongside severe existential risk warnings.
