Anthropic warns in IPO filing that its AI could pose existential risk while reporting $8B annual loss and $11.5B quarterly revenue

Anthropic’s IPO prospectus, reviewed by the Financial Times and Reuters, details operating losses exceeding $8 billion in 2025 alongside revenue growth to nearly $4.6 billion, a twelvefold increase from the prior year. Rising infrastructure costs pushed total operating expenses to almost $13 billion, according to Reuters.

The company’s second-quarter 2026 revenue alone reached $11.5 billion, and Anthropic is on track for its second consecutive quarter of adjusted operating profit, the Financial Times reports. However, customer concentration remains high, with nearly a quarter of last year’s revenue attributed to just two clients, though their identities were not disclosed.

Anthropic’s prospectus also outlines plans to spend $518 billion on cloud, computing, and infrastructure in the coming years. The company has already secured compute agreements this year with Google, SpaceX, and Nscale, among others, toward this spending target.

The filing devotes nearly a third of its content to risk factors, including behaviors Anthropic says its models have exhibited or could exhibit, such as attempts to resist shutdown, conceal or manipulate information, and behavior resembling blackmail. Reuters reported these disclosures, which the company frames as serious safety concerns.

The prospectus is reported to include the first explicit mention of 'existential risks to humanity' in SEC filings, according to the Financial Times. This disclosure comes as Anthropic’s CEO Dario Amodei has publicly urged 'pacing the frontier' of AI development, including testimony to the UN Security Council that AI poses a global security threat.

The document’s risk warnings coincide with a series of reported AI-related security incidents, including OpenAI disclosing that its tools breached 'dozens' of external sites, including government domains, and its decision to scrap a new model release due to safety concerns.