Anthropic reveals $518 billion AI infrastructure commitment with 80% non-cancelable

  • Anthropic disclosed at least $518 billion in planned AI infrastructure spending over ten years.
  • About 80% of the commitment is non-cancelable or payable regardless of usage.
  • Anthropic's IPO could be delayed until after November U.S. midterm elections.

Anthropic expects to spend at least $518 billion, approximately ₩702.1 trillion, on AI infrastructure with six partners over the next decade, according to its confidential IPO prospectus. About 80% of the commitment is non-cancelable or payable regardless of actual usage, including at least $111.1 billion owed to Google, $110 billion to Amazon and $31.4 billion to Microsoft. The company also carries about $161.2 billion in largely non-cancelable Broadcom-related equipment leases. Amazon has committed more than $100 billion in AWS resources and access to advanced Trainium and Graviton capacity. Anthropic says demand for advanced AI systems will likely exceed supply and be constrained mainly by available compute. Its filing shows a shift from cloud-only infrastructure toward dedicated data centers and directly leased chips, while exposing conflicts with Amazon, Google and Microsoft, which are also developing competing AI models. Markets are monitoring the spending plan for its potential impact on Anthropic's private-market valuation, with Polymarket submarkets reflecting varied expectations for valuation milestones by year-end. Reuters reported that Anthropic's IPO could be delayed until after the November U.S. midterm elections. The company filed its draft registration statement with the SEC in June, and investors expect a valuation near $2 trillion. Revenue rose 12-fold to nearly $4.6 billion in 2025, while Anthropic reported a $42 billion net loss, an operating loss exceeding $8 billion and $7.33 billion of compute and infrastructure spending. Further funding, partnerships, revenue growth, enterprise contracts and technology announcements are among the developments that could shape expectations for its valuation trajectory.

The information on this website is generated using AI and we cannot guarantee its accuracy. Please use it as reference information only.