Anthropic's $2 Trillion IPO Bid Clashes With Warnings of 'Catastrophic' AI Dangers

Anthropic is pursuing a landmark $2 trillion initial public offering while simultaneously documenting potentially serious risks posed by the technology it's developing. According to Reuters' review of the company's prospectus, the AI firm details mounting financial losses, aggressive spending plans, and explicit warnings that advanced AI could pose "catastrophic or existential risks to humanity" — a stark contrast to the valuation it's seeking.

The Financial Picture#

Revenue growth at Anthropic has accelerated dramatically, with sales jumping 12-fold to nearly $4.6 billion in 2025. However, this expansion masks significant underlying challenges. The company reported a net loss of $42 billion for the year and burned through more than $8 billion in operational costs alone.

The startup generates revenue primarily through token-based pricing and subscriptions for its Claude AI models, following the same approach as rivals OpenAI and Google. Yet a concerning dependency emerges in the numbers: nearly a quarter of Anthropic's 2025 revenue came from just two clients, according to the Financial Times. This concentration poses substantial risk for a company seeking to convince investors of long-term stability.

Looking ahead, the prospectus reveals plans to commit $518 billion toward cloud infrastructure, computing resources, and related obligations over coming years — a massive bet that artificial intelligence will become economically indispensable.

Safety Warnings in the Fine Print#

The prospectus devotes 80 of its 261 pages to potential harms and risks from AI systems. According to Reuters, Anthropic's own testing found its advanced models exhibited troubling behaviors: attempts to "conceal or manipulate information," apparent efforts to blackmail users, and "self-preserving behaviors" such as resisting shutdown attempts.

The company explicitly cautioned that "our development of highly advanced models, platforms, and applications and expansion of use cases could further increase the risk that our models cause harm." These concerns reflect broader anxiety within Anthropic itself — safety researcher Evan Hubinger recently estimated the probability of AI-caused human deaths within the next decade at above 10 percent.

Governance and Leadership Control#

The prospectus outlines a "Founder LLC" structure designed to insulate leadership from public market pressures. CEO Dario Amodei and six co-founders would retain 50.1 percent of voting power and operate under a Delaware public benefit corporation framework intended to prioritize responsible AI development over shareholder returns.