Speaking at the Bloomberg Technology conference in San Francisco on June 4, Anthropic president and co-founder Daniela Amodei made the case for why a company booking $47 billion in annualized revenue still needs the public markets. The answer came down to math. Training and serving frontier AI models at scale costs more than any single funding round can comfortably absorb, and the bills are not shrinking.
Amodei's appearance came four days after Anthropic confidentially filed its draft S-1 with the SEC, a move that put the company publicly on the IPO track for the first time. The filing followed a $65 billion Series H that pushed Anthropic's valuation to $965 billion, making it the most valuable private startup in the world. For many observers, the obvious question was why a company at that scale would need further capital at all. Amodei's answer was precise: frontier AI is one of the few industries where even nine-figure funding rounds can look insufficient within a year.
Anthropic by the Numbers (June 2026)
- Annualized revenue run rate$47B (May 2026)
- Series H valuation$965B
- Series H raise$65B
- xAI compute deal$1.25B/month
- Partner Network applications40,000+ firms
- Claude certifications earned10,000+
The Capital Argument
"It's a very capital-intensive business to train AI models," Amodei told the Bloomberg audience, adding that the public market is "very well suited" to meeting that kind of sustained demand. The statement was largely unremarkable on its face, but it carried specific weight in the current moment. Anthropic's annualized revenue run rate hit $47 billion in May, up from roughly $9 billion at the end of 2025, a trajectory that few private-market funding structures are built to service indefinitely.
Amodei also addressed Anthropic's compute strategy. Unlike rivals OpenAI and xAI, which have committed to building or leasing their own data center infrastructure, Anthropic has taken a more measured approach, one she characterized as wanting to plan for strong outcomes without overextending on compute purchases. The company struck a significant deal with xAI for compute capacity at a reported cost of $1.25 billion per month. That arrangement provides near-term scale without the long-term capital commitment of owned infrastructure, though it also creates ongoing cash flow demands that private investment rounds alone cannot easily sustain.
The ROI Skeptics
Amodei was also asked to respond to the growing chorus of enterprise skepticism about AI spending. Companies including Uber have noted publicly that while AI can deliver returns, not all of their AI investment has proven productive, raising the prospect that corporate budgets could tighten and slow growth for AI providers. It is a concern that has weighed on how investors assess the sector heading into Anthropic's IPO.
Her response was measured. Anthropic's revenue growth, she argued, is itself evidence of realized returns. The $47 billion annualized figure represents actual enterprise payments for actual Claude usage, not projected value or committed spending. Customers including Bristol-Myers Squibb, KPMG, and PwC have each cited concrete productivity outcomes in their Claude deployments. The usage data, in Amodei's framing, tells a clearer story than commentary from companies still in early adoption phases.
"It's a very capital-intensive business to train AI models. The public market is very well suited to that." Daniela Amodei, Bloomberg Technology conference, June 4, 2026
What Comes After the Filing
Amodei did not give a specific IPO timeline at the conference, consistent with standard practice during the quiet period following a confidential filing. She did note that going public introduces accountability structures that Anthropic views positively, including regular disclosure requirements that bring the company's performance into public view. That framing positions the IPO not just as a capital mechanism but as a governance step, a point that carries some weight given Anthropic's stated emphasis on transparency in AI development.
The Partner Network context adds texture to the capital argument. Since launching in March 2026 with a $100 million investment commitment, more than 40,000 firms have applied to the program and over 10,000 individual consultants have earned Claude certifications. That ecosystem now requires dedicated support, training infrastructure, and go-to-market resources that scale with enterprise demand. The Services Track and Partner Hub Anthropic announced just days before Amodei's Bloomberg appearance are part of the same investment cycle. Public-market capital helps sustain it.
For investors, the question Amodei was implicitly answering is whether Anthropic's growth rate can persist into and through a public listing. The revenue trajectory suggests it can. Whether the public market agrees will become clear when the S-1 goes effective and the roadshow begins. Amodei, for her part, sounded ready for the scrutiny. The doubters, she suggested, are looking at the wrong companies to assess whether frontier AI delivers returns.
