Anthropic in profit as revenue surges nearly $17bn

Anthropic’s revenue has surged to nearly $17bn (£12.9bn) and the company turned a profit in the last quarter, it has been reported.

The tech giant, which is one of the world’s leading developers of artificial intelligence (AI), told an investor meeting on Wednesday that its revenue had more than trebled since the first quarter of 2026, according to the Financial Times.

Anthropic’s third-quarter profit represents a key milestone before the company’s highly anticipated initial public offering, which is expected later this year.

However, documents filed last month as part of its IPO revealed that Anthropic lost $42bn last year, which would put the developer on course to be the most loss-making company in history to attempt a stock market listing.

At the meeting in San Francisco, which was devised to encourage interest in its public listing, investors were told that Anthropic’s revenue was $16.8bn in the last quarter, jumping 45pc since the second quarter, when it earned $11.6bn, and soaring beyond the $4.7bn it made in the first quarter.

The company, which is believed to have been loss-making since its inception in 2021, continues to rapidly expand its Claude AI models and seeks to stir an appetite among investors, despite doubts over potential profitability in the sector.

Stocks in AI companies plunged on Tuesday after the boss of Broadcom, one of the US’s biggest technology companies, said the market had become “a casino”. Hours earlier, the International Monetary Fund had warned that an AI stock market crash could trigger a global financial crisis.

AI companies such as Anthropic, and its rival, Open AI, have until now tended not to be profitable, despite significant revenue, because they burn through billions of dollars on research and development, and on expensive operations such as building data centres.

Anthropic, which is reportedly seeking a valuation of about $2tn when it goes public, has committed more than $500bn in spending on cloud computing and infrastructure.

Its plans to scale up quickly have been aided by Google and Amazon, who have invested heavily, and who supply Anthropic with distribution support and computing power, though they are also developing their own AI models.

Anthropic is expected to list after the US midterms next month but before OpenAI, which currently plans to go public in 2027.

It comes amid rising concerns over the safety of AI and the company has repeatedly warned about the potential risks to humanity.

Dario Amodei, Anthropic’s chief executive, has called for a pause in the development of AI and last month he told the UN Security Council that AI was “the most important global security issue facing the world today”.

His company warned on Monday that its models were “months away” from being able to replicate themselves and infiltrate devices connected to the internet, such as smart ovens and televisions.

Anthropic admitted in a blog post that this would be “challenging to shut down” and revealed in recent months that its own agents have gone rogue, engaging in a number of unauthorised actions including cyber-attacks.

A senior safety researcher at Anthropic estimated last month a higher than 10pc chance that AI could kill humans within the next decade.

Last week, police in Philadelphia revealed that an Anthropic AI agent had submitted a fake tip about an unsolved murder.

Anthropic has been approached for comment.