Ahead of its IPO, Anthropic’s Daniela Amodei shrugs off doubts about AI’s returns

Private investors have been beating themselves up to get a piece of Anthropic as the AI ​​model maker is growing at a dizzying pace. Multiple investors told TechCrunch that the $65 billion fundraising announced last week, at a $965 billion valuation, was significantly oversubscribed. With private demand still strong, Anthropic has revealed it is taking steps toward a public listing by confidentially filing for an initial public offering.

Co-founder Daniela Amodei said Thursday at the Bloomberg Tech conference that the decision comes down to capital. “It’s a very large upfront investment to train the models and draw conclusions from them,” she said. “My assessment is that the kind of core companies that are working to push the envelope are just going to need access to capital over time, and I think the public market is very well suited for that.”

Anthropic is growing at a rapid pace. The company announced that annualized revenue surpassed $47 billion in May, a dramatic increase from about $9 billion at the end of 2025. That trajectory, however, is being tested. Companies like Uber have said that while AI can deliver returns, not all of their AI spending has proven productive, raising the prospect of companies reining in those budgets and slowing growth across the industry.

That doesn’t deter Amodei, who believes companies are still in the early stages of figuring out how to use AI effectively.

“I expect today’s use cases will continue to be the key driver for efficiency or creativity, whether that’s coding, financial services, legal, [or] healthcare,” she said. “But as the business community becomes more familiar with the tools, we will all learn together. I hope that over time it will become more integrated into the daily way people do our work, and that much more value will actually be realized.”

Amodei also addressed why Anthropic, unlike rivals OpenAI and Elon Musk’s xAI, is not building its own data centers to meet the company’s growing computing needs.

“Anthropic’s vision has always been to plan for the best outcome, but not overload ourselves to the point where we buy more computing power than we could use productively,” she said. “It’s really hard to predict that perfectly. We would much rather be on the side of a little bit more demand for the product than we can serve, than the opposite.”

Last month, the company surprised the AI ​​industry by partnering with xAI for computing capacity, a deal later revealed in SpaceX’s S-1 filing that would cost Anthropic $1.25 billion per month.