At Anthropic, the race to build cutting-edge artificial intelligence (AI) is increasingly defined by one constraint: computing, and at the center of the company's response is its low-profile chief financial officer, Krishna Rao, who has emerged as a key architect of its rapid expansion.
In February, Anthropic raised US$30 billion in a Series G round led by GIC and Coatue Management, valuing the company at US$380 billion post-money. The round drew a broad coalition of global investors, including D. E. Shaw Ventures, Dragoneer Investment Group, Founders Fund, ICONIQ Capital, and MGX, alongside participation from tech giants like Microsoft and Nvidia.
The funding is being directed toward frontier research, product development, and most critically, infrastructure, as Anthropic seeks to maintain its position in enterprise AI and coding tools. "The message from our customers is the same: Claude is increasingly becoming critical to how businesses work," Rao said, pointing to demand from both startups and large corporations.
That demand has translated into extraordinary growth. Less than three years after generating its first dollar of revenue, Anthropic has reached a US$14 billion run rate, expanding more than tenfold annually over that period. More recently, that figure has surged past US$30 billion, driven by the rapid adoption of its Claude models.
To sustain that trajectory, Anthropic is making its largest infrastructure bet yet. The company has signed a new agreement with Google and Broadcom to secure multiple gigawatts of next-generation TPU capacity, expected to come online starting in 2027. Most of that compute will be based in the US, expanding on an earlier US$50 billion commitment to domestic infrastructure.
The deal reflects a strategy Rao has championed since joining in 2024: diversify supply. Anthropic now operates across chips from Nvidia, Google, and Amazon, and remains the only frontier AI developer with models deployed across all three major cloud platforms, including Amazon Web Services (AWS), Google Cloud, and Microsoft Azure.
That flexibility has become a competitive advantage in an industry constrained by scarce chips and soaring demand. But it also underscores the scale of the challenge. Training and running advanced AI models requires vast, costly clusters of hardware, and competition for those resources is intensifying — particularly with rivals like OpenAI.
Rao, who previously held roles at Blackstone and Airbnb, is now tasked with balancing that growth against financial discipline. He has already overseen roughly US$60 billion in fundraising, helping position Anthropic among the most valuable private companies in the world.
Yet even as capital pours in, the margin for error remains thin. Anthropic's chief executive, Dario Amodei, has warned that misjudging future compute needs could determine whether the company succeeds or fails.
For Rao, the job is no longer just about raising money. It is about securing the physical backbone of the AI era and doing so fast enough to keep up with demand that shows no sign of slowing.
Article edited by Jack Wu
