AI giant Anthropic plans to sell shares in US as valuation nears $1tn

Artificial intelligence (AI) firm Anthropic has revealed its plans to go public in the US, a move which would allow people to buy and sell its shares on the stock market.

The company behind chatbot Claude said on Monday it had filed paperwork with the US authorities in order to make an initial public offering (IPO) this year.

Anthropic said the price and number of shares to be offered "have not yet been set".

The firm's stock market plans, coming alongside those of Elon Musk's SpaceX , is set to show whether investor appetite matches the soaring valuations of AI companies.

Anthropic, founded just five years ago, recently raised money from private investors that valued the company at more than $965bn (£717bn), ahead of OpenAI's most recent valuation of $852bn.

Anthropic chief executive Dario Amodei created the firm after leaving OpenAI, where he had worked for several years, over disagreements with its chief executive Sam Altman.

The two firms have since become fierce rivals in the AI world, developing similar technology and fighting for the attention and spending of users and corporate customers.

OpenAI is also reportedly considering going public this year.

Altman told CNBC on Monday that, while his company did intend to go public, it was in no rush to do so.

"We'll do it when it makes sense," Altman said.

Should OpenAI's listing happen, the US capital markets are poised to see an historic level of investment in just a handful of companies.

SpaceX alone is expected to break stock market records with its target valuation, but the potential value of Anthropic and OpenAI are not far behind.

Meanwhile, Google's owner Alphabet has revealed plans to raise $80bn to spend on AI.

Matt Britzman, senior equity analyst at Hargreaves Lansdown, said the announcement was "a clear sign that the AI arms race is moving into a more capital-hungry phase".

SpaceX files for stock market debut that could make Elon Musk a trillionaire

Listing will be 'yardstick for investors'

Before firms list, they have to publish an IPO prospectus, which details the company's finances, management and risks.

Tineke Frikkee, senior fund manager at W1M, said she would be interested to study the prospectus to find out what profit Anthropic is making.

"No doubt it'll be a multi-hundred pages document, but that's what we'll be looking for," she told the BBC's Today programme.

Sana Kharegani, chief strategy officer of AI firm Era 4, said Anthropic's decision to go before OpenAI's potential listing "carries a bit of a risk".

"What we are going to get is a precedent – a gauntlet laid down for metrics that might matter. So maybe enterprise revenue, or subscriber counts, or hard numbers rather than philosophy," she told the BBC.

"But if things don't go exactly as planned... you're leaving the door for others to come after you," she added.

However, Troy Hooper, a leader of equity capital markets at Mergermarket, said going second, not first, was the risky move.

"The first mover has a real chance to define how public markets value generative AI, setting up the yardstick that investors will use to measure everyone else," he said.

"Neither Anthropic nor OpenAI wants to be the last major AI pure-play to list."

Harrison Rolfes, a research analyst a Pitchbook, added Anthropic's IPO "will be the most scrutinized public offering in tech history", with investors poring over its business margins, sales and profitability for signs that valuations and costs of AI make financial sense.

By being so close to SpaceX's upcoming stock market debut, Rolfes said that the two companies "represent the largest concentration of pre-IPO capital ever brought to market simultaneously".

"The 2026 window either becomes the most consequential IPO cycle since the dot-com era or the most expensive lesson in narrative-versus-fundamentals that public markets have ever taught," Rolfes said.

Anthropic has been fighting this year , too, with the US Department of Defense (DoD).

After the DoD late last year insisted that Anthropic should accept contractual terms as part of a $200m deal, which stated government agencies could use AI tools like Claude for "any lawful use", Amodei made his concerns public.

Although Claude was the first modern AI chatbot and model to be deployed on US government networks that were classified, the new contract language suggested to Anthropic the possibility of its AI being used in mass domestic surveillance or on fully-autonomous weapons of war.

President Donald Trump said the US would "never do business with [Anthropic] again" and Defence Secretary Pete Hegseth banned any US agency from using Claude.

Anthropic then launched legal action against the government. Although there are recent signs from the White House that tensions with Anthropic have cooled, the lawsuit is ongoing.

The split with the Trump administration does not appear to have scared off other Anthropic customers.

The company has told investors that it expects to turn a profit in the first half of this year, as it says sales of its Claude product and related services have grown significantly.

Neither SpaceX nor OpenAI are currently profitable.