By PYMNTS | October 15, 2026
Anthropic has reportedly told investors that its third-quarter revenue had jumped to $16.8 billion as it prepares to go public.
That figure is up from $4.8 billion in the first quarter and $11.6 billion in the second, the Financial Times (FT) reported late Wednesday (Oct. 14), citing sources familiar with the matter.
The report added that Anthropic turned a profit in the third quarter, and notes that this is happening as Anthropic is trying to drum up support for a massive initial public offering amid concerns about the AI sector. The company’s growth is slower than earlier this year and below some of its backers’ expectations, investors and analysts told the news outlet.
Anthropic convened an investor meeting at its office in San Francisco Wednesday for a full day of presentations, with executives arguing that a safety-focused philosophy, speedy growth and improved efficiency would set the stage for its IPO in November.
The company told investors it was profitable on an adjusted basis in the third quarter for the second straight quarter, multiple sources with knowledge of the matter told the FT.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
While the company did not share an update on its annualized revenue, the FT cites figures from research group SemiAnalysis showing projected sales increasing to roughly $74 billion on an annualized basis.
Those estimates are lower than more enthusiastic predictions during the summer, when a number of Anthropic’s investors told the FT the company would far exceed $100 billion in annualized revenue by year-end.
A Wall Street Journal report on Wednesday’s meeting said that investors expect the startup to reach $110 billion in annualized revenue by the end of the year.
Joey Brookhart, an analyst who covers AI labs for SemiAnalysis, told the FT investors had been downgrading their projections for Anthropic’s annualized revenue over the last three months. The report added that expectations have dipped as major customers like Meta increasingly use their own AI tools.
In other AI news, recent PYMNTS Intelligence research shows that 89% of card issuers plan to adopt or expand at least one fraud capability powered by the technology. Half of them point to fraud detection and scoring as the single most crucial AI capability to add in the next year, while 63% plan real-time fraud detection specifically.
“Fraud decisions happen continuously, against live transactions, and often have to be made quickly enough to approve a legitimate purchase without creating unnecessary friction for the cardholder,” PYMNTS wrote Wednesday.
“Issuers are putting AI into that environment at the same time that they are shortening their technology investment cycles.”
The research found that the average time until an issuer’s next platform upgrade has dropped from 7.8 years to 2.4 years, with 70% of issuers planning an upgrade in the next three years.