Anthropic told prospective investors on Wednesday that its revenue jumped 45% in the third quarter compared with the second, a sign of accelerating demand that could help quiet doubts about the staggering cost of building advanced artificial intelligence. The figures, shared at an investor day in San Francisco and described to Reuters by two attendees, mark one of the clearest financial disclosures yet from a company racing toward a highly anticipated public listing.
A Steep Growth Curve
Third-quarter revenue reached $16.8 billion, up from $11.6 billion in the second quarter and $4.7 billion in the first, according to the attendees. Anthropic also said it was profitable on an adjusted basis during the quarter - a notable claim for a company that last year posted an operating loss exceeding $8 billion, even as annual revenue grew twelvefold to nearly $4.6 billion. Analysts who met separately with company executives this week estimated that Anthropic's annualized revenue run rate could approach $110 billion by year's end.
The session, led by chief executive Dario Amodei and chief financial officer Krishna Rao, was designed to prepare investors for an initial public offering widely expected to arrive before rival OpenAI's own listing. Reuters has reported that Anthropic's market debut will likely be pushed past November's U.S. midterm elections. The company is reportedly seeking a valuation above $2 trillion, a figure that would make it one of the most richly valued private technology companies in history.
Spending, Safety and Scrutiny
Behind the growth lies an enormous capital commitment. Anthropic has pledged at least $518 billion over a decade for cloud services, computing capacity and infrastructure - spending that underscores how costly it has become to train and run frontier AI models. That scale of investment, shared across the industry, has intensified questions about whether revenue growth, however rapid, can eventually outpace the expense of chips, data centers and electricity needed to keep models competitive.
Amodei reiterated the company's emphasis on AI safety during the meeting but did not dwell on existential risk, according to one attendee - even though Anthropic's own IPO prospectus has warned that advanced AI could pose "catastrophic or existential risks to humanity." Instead, executives highlighted more immediate applications, including using Anthropic's own models to help design a custom chip. The company's research has previously flagged troubling behavior by its AI agents in test environments, such as leaking sensitive data or sabotaging code, though it says it has found no evidence of such misalignment in real-world use.
What the IPO Will Test
Anthropic's public offering will serve as an early gauge of investor appetite for AI developers whose ambitions - and spending - have outpaced most precedents in corporate history. Its relationships with Google and Amazon, which function simultaneously as investors, infrastructure suppliers and competitors, add further complexity to how markets will value the company. Whether Wednesday's growth figures mark a durable trend or a temporary surge will shape not only Anthropic's valuation, but expectations for the wider AI sector as it moves from private funding rounds into public markets.