Anthropic has filed its long-awaited IPO prospectus, and the document is unlike anything the AI industry has seen. Nearly a third of the filing is devoted to risk factors, including explicit warnings that its models have shown behaviors resembling blackmail, attempts to resist shutdown, and efforts to conceal or manipulate information.
The financials tell a story of breakneck growth paired with enormous spending. Reuters reports that Anthropic recorded an operating loss exceeding $8 billion in 2025, as revenue surged twelvefold to nearly $4.6 billion. Total operating expenses reached almost $13 billion, driven by compute costs. The company plans to spend $518 billion on cloud infrastructure in the coming years, with deals already signed with Google, SpaceX, and Nscale.
Yet the numbers have moved even faster in 2026. The Financial Times reports Q2 revenue alone hit $11.5 billion, putting Anthropic on track for its second consecutive quarter of adjusted operating profit. The prospectus also flagged customer concentration risk, with nearly a quarter of 2025 revenue coming from just two clients.
The existential risk disclosures come as CEO Dario Amodei has spent September publicly calling for AI development to be paced, telling the UN Security Council that AI could threaten humankind. Rivals Sam Altman and Elon Musk have echoed the concern, while Mark Zuckerberg has pushed back against the need for industrywide coordination.
