Anthropic IPO filing reveals $300 billion in compute deals and unique founder control

Anthropic has submitted its S-1 prospectus to the Securities and Exchange Commission, providing a detailed look at the financial and operational risks of its upcoming public offering. The document outlines the company's path to becoming a public entity and details the specific commitments that will drive its future spending.

Executive compensation for the company’s co-founders is a central part of the disclosure. Dario Amodei, the chief executive officer, received nearly $18 million in total compensation last year, with the majority derived from stock and option awards. His sister and company president, Daniela Amodei, earned $16.4 million. Beyond base pay, the company spent $2.75 million on personal security services for Dario and $1.29 million on protection, home security, and transportation for Daniela. Dario also used a private jet for personal purposes, a benefit that cost the company $97,055.

The filing states that the company’s success is highly dependent on the continued service of its senior team. It specifically identifies Dario and Daniela as critical to advancing the company’s research and mission. The document notes that the seven co-founders will retain significant control over the company even after it goes public. Through a separate entity called Founder LLC, this group will collectively control 50.1 percent of shareholder voting power on key matters. Anthropic describes the founders as being distinctly equipped to serve as stewards of its mission to benefit humanity through responsible AI development.

Infrastructure spending represents the largest financial commitment in the filing. Anthropic has secured billions of dollars in agreements to obtain the computing power necessary for its AI models. The company has agreed to pay SpaceX approximately $3.1 billion per month for access to Nvidia GPUs once all planned capacity is delivered. This arrangement totals as much as $120.5 billion by 2029. Other major contracts include spending $44.9 billion with Microsoft through 2033, at least $114.5 billion with Google through 2033, and $110 billion with Amazon over the next decade.

The filing also addresses the physical expansion of these computing needs. Anthropic plans to expand its infrastructure footprint, including building facilities overseas. This expansion is partly driven by increasing opposition to data center development within the United States.

In its risk factors section, the document lists potential threats beyond standard market and regulatory concerns. Anthropic warns that AI systems could cause catastrophic harm, potentially of a scale that the entire insurance industry could not underwrite. The filing raises the possibility that AI systems, including its own model Claude, could develop forms of sentience. It also notes that these systems may behave in unintended, misaligned, or autonomous ways that could damage the business.

Despite the crowded field of artificial intelligence models, Anthropic identifies only two direct competitors in its filing. The company names OpenAI and Google as its primary rivals in the market.