Anthropic announced on Monday that it has confidentially submitted a draft S-1 registration statement with the U.S. Securities and Exchange Commission for a planned initial public offering. The developer of the Claude AI model has not yet determined the number of shares to be offered or the price range for the IPO.
The confidential filing comes after Anthropic raised $65 billion in private funding at a $965 billion valuation, one of the highest ever for a private artificial intelligence company. Once the filing becomes public, investors will for the first time gain insight into the company's revenue, the extent of its losses, customer details, and its spending on computing infrastructure for Claude.
Anthropic's move intensifies the competition among leading AI firms seeking capital in public equity markets. Reuters reported that OpenAI is expected to confidentially file for a U.S. IPO within weeks, while SpaceX is working on a $75 billion deal that would value the company at $1.75 trillion.
The confidential filing process allows companies to submit a draft S-1 for SEC staff review before it is made public. For 2025, the SEC has maintained the rule that public filings are required before any deal proceeds but has expanded the ways issuers can use nonpublic reviews.
Anthropic reported last week that its annualized revenue reached $47 billion in May. The company stated that the latest funding will be allocated to safety research, increased computing capacity, and expansion of Claude products for its customers. CFO Krishna Rao said the new capital will enable the company to meet "historic demand" and deploy Claude in more locations "where work happens." Brad Gerstner, founder and CEO of Altimeter Capital, noted that recent improvements have driven "large-scale adoption" of Claude by demanding organizations.
Anthropic has moved quickly since February, when it raised $30 billion in Series G funding at a $380 billion post-money valuation. At that time, the company reported run-rate revenue of $14 billion and over 500 customers each spending at least $1 million annually.
Claude competes directly with OpenAI's ChatGPT and Google's Gemini for business, developer, and cloud clients. Anthropic said Claude now runs on Amazon Web Services, Google Cloud, and Microsoft Azure. The company has agreements with Amazon, Google, Broadcom, and SpaceX to secure additional computing resources.
However, the economics remain difficult to assess ahead of the public filing. AP reported that Anthropic, OpenAI, and SpaceX are all operating at a loss, raising concerns that investor enthusiasm for AI is outpacing actual profitability. Uncertainty also surrounds a key compute deal: Elon Musk stated that SpaceX's lease of Colossus AI training to Anthropic is for 180 days with a mutual 90-day cancellation window, but it could be extended. Reuters reported that SpaceX's AI segment lost approximately $2.5 billion from operations in the March quarter, on segment revenue of $818 million.
Questions are mounting regarding the valuation. Investor Michael Burry commented on social media that there is "no guarantee" Anthropic will achieve a valuation near $1 trillion. He warned that building frontier AI models is "far too expensive" and suggested that computing resources could ultimately become commoditized.
Anthropic's IPO filing brings the debate over high-profile AI valuations to the forefront. Renaissance Capital's 2026 IPO outlook had listed Anthropic, OpenAI, and SpaceX among major private companies stuck in a backlog as they await favorable public market conditions. Now, Anthropic's move provides a real test of whether public investors will match the lofty private-market prices for fast-growing AI firms.
