Anthropic Seeks About $60B in Chip Financing, Offering Blackstone Over 10% Yield for Compute Buildout

Anthropic, the creator of the popular AI chatbot Claude, is in discussions to secure approximately $60 billion in financing for its data center infrastructure, according to the Financial Times. This massive financing package would involve Google providing around 1.5 gigawatts of custom Tensor Processing Unit (TPU) capacity. Blackstone, a prominent private equity firm, is among the funds considering a $9 billion investment in subordinated debt, with potential returns exceeding 10%.

The proposed financing rate is significantly higher than the 8.5% rate seen in financing deals just four months prior, signaling a shift towards more costly capital for AI infrastructure. This marked increase in financing costs underscores the closing of the era of cheap, abundant capital for advancing AI technologies. Analysts interpret this development as a significant structural change in the industry, where the focus is shifting from raising capital at the highest valuations to financing compute at the lowest possible cost.

In this evolving landscape, private credit funds like Blackstone are stepping in to fill the gap left by more selective venture equity. As AI labs pursue alternative financing methods to fund their data center buildouts, Anthropic's move aligns with a broader trend of AI companies leveraging future revenue streams to finance their capital expenditures.

This strategic shift comes as Anthropic prepares for a potential record-setting IPO, with Q3 revenue reported at $16.8 billion. The competitive landscape for compute resources is becoming increasingly challenging for AI startups and specialized firms. If Anthropic manages to secure financing at double-digit yields while its more financially robust competitors fund compute at internal hurdle rates of 6-8%, the financial disparity could widen substantially over time.

The successful execution of Anthropic's financing strategy will shed light on whether private credit can reliably bridge the gap between AI revenue growth and the substantial upfront costs of compute infrastructure. This deal could mark the beginning of a new AI credit cycle, reshaping the dynamics of the rapidly evolving AI technology sector.