Anthropic files for US stock market debut after valuation surge

Analysts expect Anthropic to make its public market debut as early as the autumn

AI firm Anthropic has confidentially filed for an initial public offering (IPO) in the United States.

Anthropic, which develops the Claude family of AI assistants, announced the filing on Monday but did not disclose the size of the offering, the number of shares it intends to sell, or the valuation it hopes to achieve when it begins trading.

The move positions Anthropic among a wave of high-profile technology firms seeking to capitalise on investor enthusiasm for AI, with rivals including OpenAI and Elon Musk's SpaceX also expected to pursue public listings this year.

Anthropic's confidential filing comes only days after the company completed a funding round that valued the business at $965bn, nearly tripling its valuation from $380bn in February.

The latest fundraising has propelled the company ahead of OpenAI to become the world's most valuable privately held startup.

Anthropic said the IPO would proceed after a review by the US Securities and Exchange Commission (SEC) and would depend on market conditions and other factors.

Industry analysts expect Anthropic could make its public market debut as early as the autumn.

Rapid rise

Founded by former OpenAI researchers, Anthropic was once viewed as a secondary player in the race to build advanced AI systems.

However, its fortunes have changed dramatically over the past year, driven in large part by the success of tools such as Claude Code and its workplace-focused AI offerings, which have gained traction among enterprise customers seeking to automate programming and productivity tasks.

"Claude is increasingly indispensable to our growing global community of customers," Anthropic Chief Financial Officer Krishna Rao said in a statement.

"This funding will help us serve the historic demand we are experiencing, stay at the research frontier, and bring Claude to more of the places where work happens."

Google to raise up to $80bn

Anthropic's public market ambitions come amid an escalation in AI spending.

Google parent company Alphabet announced plans this week to raise up to $80bn through a share sale to fund a massive expansion of its AI infrastructure. The fundraising effort includes a $10bn investment from Berkshire Hathaway.

Alphabet's shares fell 4% in early trading on Tuesday following the announcement as investors weighed the scale of the company's spending plans.

The technology giant now expects capital expenditure of between $180bn and $190bn in 2026, roughly double projected spending levels for 2025, with further increases anticipated the following year.

The company says the funds will be used to build new datacentres, develop custom AI chips and expand the computing infrastructure supporting its Search, Cloud and Gemini businesses.

"AI is driving an expansionary moment for Alphabet," the company said, adding that customer demand for AI services was exceeding available computing capacity.

Anthropic-Google partnership

Anthropic and Google have strengthened their partnership in recent months through a series of major AI infrastructure and cloud computing agreements.

Last month, reports indicated Anthropic had agreed to spend approximately $200bn on Google Cloud services over five years.

The companies have also expanded cooperation around Google's tensor processing units (TPUs), with Anthropic securing access to multiple gigawatts of future AI computing capacity expected to begin coming online from 2027.

Bloomberg previously reported that Google was considering investments of up to $40bn in Anthropic, tied to performance milestones and the startup's continued growth.

Computing says:

The timing of the Anthropic filing is interesting. Anthropic let it be known recently that it projects an operating profit of $559 million for Q2. However, Anthropic chose not to share whether the remaining two quarters of the year will be equally as profitable – or profitable at all.

As blogger and podcaster Ed Zitron recently reported, Anthropic currently receives a discount on its compute costs as part of the terms of its deal to take over the SpaceX owned Colossus 1 datacentre in Memphis. We don’t know how much but of a discount, but the SpaceX IPO filing says “capacity ramping in May and June 26 at a reduced fee”.

Elon Musk also recently publicly walked back earlier reporting of the deal as a $1.25bn-per-month, three-year arrangement running through May 2029. It is, according to Elon Musk a 180-day base lease with a mutual 90-day cancellation right on either side after that.

Anthropic’s move to token-based billing in Q1 is already having a serious impact on AI costs for companies such as Amazon, Uber and Walmart. It would be a reversal of the usual laws of economics for the sharp increase in cost of LLM coding tools not to reduce demand – and Anthropic’s revenues and profits.

To be clear, Anthropic hasn’t done anything that looks dodgy. But the selective leaking of information pertaining to one quarter only should maybe raise some eyebrows among private and institutional investors.