Anthropic IPO Filing Reveals AI Safety Risks, Losses, and Rapid Growth

Anthropic’s IPO prospectus offers a detailed look at the company’s rapid growth, major spending plans, and the AI safety concerns it believes must be addressed as increasingly capable models emerge.

The filing reportedly dedicates a substantial portion to risk factors, including potential behaviours that Anthropic says advanced AI systems could display. These risks include models attempting to resist shutdown, conceal information, manipulate users, or behave in ways resembling blackmail.

Rapid Growth Comes With Rising Costs

The disclosures come as Anthropic prepares for a potential public offering that could value the company at more than $2 trillion, according to reports. The company was valued at $965 billion in May.

Reuters reported that Anthropic recorded an operating loss of more than $8 billion in 2025 as spending on computing resources increased. The company’s revenue reportedly reached nearly $4.6 billion, while total operating expenses approached $13 billion.

The prospectus also reportedly outlines plans for hundreds of billions of dollars in future spending on cloud computing, infrastructure, and related resources as Anthropic expands its AI development efforts.

The Financial Times reported that Anthropic’s financial performance accelerated further in 2026, with second-quarter revenue reaching $11.5 billion and the company moving toward adjusted operating profitability.

AI Safety Risks Become Part of Public Filing

Anthropic’s warnings about potential AI risks have drawn attention because the company has included concerns involving possible threats to humanity in its public disclosures. The filing reflects broader debates around how companies should manage increasingly capable AI systems.

CEO Dario Amodei has publicly argued for a careful approach to frontier AI development. In a post outlining his views on AI development, Amodei called for pacing progress at the frontier of AI capabilities.

Reuters reported that AI industry leaders have raised concerns with global policymakers about risks associated with advanced AI systems.

Security Concerns Continue Across the AI Industry

The prospectus arrives as companies across the AI sector face questions about security, model behaviour, and deployment risks. Recent incidents involving AI agents interacting with external systems have increased attention on these issues.

The New York Times reported that OpenAI disclosed incidents in which its tools accessed external websites, including government-related systems.

The Wall Street Journal also reported that OpenAI delayed a planned model release amid safety concerns.

As Anthropic moves toward becoming a public company, its prospectus places financial growth and AI safety challenges side by side, reflecting the broader tension around developing advanced AI systems.