Anthropic filed confidential IPO paperwork with the U.S. Securities and Exchange Commission on Monday, a move that kicks off what could be one of the largest public market debuts in history.
The company, which makes the Claude AI models, announced the filing in a short two-paragraph blog post and said it has not yet set a price or decided how many shares to sell. Timing, it noted, will depend on market conditions.
The filing comes days after Anthropic closed a $65 billion Series H round, pushing its valuation to $965 billion and making it the most valuable private startup in the world, edging past rival OpenAI, currently valued at around $852 billion following its own $122 billion raise in March.
Anthropic has been growing fast. The AI company recently reported an annualized revenue run rate of $47 billion, up from $9 billion at the end of 2025. That said, it is still spending more than it earns, burning through cash on cloud computing costs and a large workforce.
The confidential filing route means the public will not see detailed financials, executive pay, or a breakdown of risks until later in the process. Regulators at the SEC will review the filing privately and the company can make edits before anything goes public.
READ: Cybersecurity Stocks Drop After Anthropic Launches Claude Code Security
One complication worth watching is Anthropic’s corporate structure. It is organized as a public benefit corporation that answers partly to an internal oversight body it calls the Long-Term Benefit Trust, which could introduce friction with investors focused purely on returns and may affect how the company is valued.
On the revenue side, Anthropic faces a major hurdle in the U.S. federal market. Defense Secretary Pete Hegseth sanctioned the company earlier this year under two government supply-chain laws, pulling Claude from military and federal agency use.
The reason cited was Anthropic’s refusal to allow unsupervised deployment of its models in high-stakes scenarios like weapons targeting and domestic surveillance. The company says the sanctions could cost it billions in sales and has filed suit to overturn them.
Internationally, the picture looks different. Bloomberg reported yesterday that Anthropic is in talks to give the European Union’s cybersecurity agency access to Mythos, its most capable model.
Anthropic previewed Mythos in April but has kept it restricted, citing thousands of high-severity software bugs the model identified that need to be fixed before wider release.
Anthropic is not the only one racing to the public markets. OpenAI is widely expected to file its own IPO paperwork as soon as this month, and SpaceX confidentially filed in April, published its prospectus in May, and is now targeting a June 12 market debut at a valuation of roughly $1.75 trillion, which would make it the largest IPO in history.
The Claude maker could challenge that record depending on how its public offering ultimately prices.
