Descriptions:
Bloomberg Technology sat down with CoreWeave CEO Michael to discuss a new multibillion-dollar compute deal with Anthropic — the AI safety company’s first contract with a provider outside its core investor group — coming just days after CoreWeave announced a separate $21 billion agreement with Meta. The interview offers a rare look at how hyperscale AI infrastructure deals are structured and financed, with CoreWeave using signed customer contracts as collateral to access debt markets at below-market rates.
On the capital side, CoreWeave disclosed it issued $8.5 billion in bonds at negative spread — a direct result of counterparty quality — and executed what it called the largest-ever dual offering combining convertible notes and high-yield paper. The CEO frames the firm’s growth model as a ‘land and expand’ strategy: initial contracts with major AI labs grow into deep infrastructure partnerships over time.
The conversation also touches on broader competitive dynamics, including OpenAI reportedly pulling back its Stargate expansion in the UK due to rising energy costs, and how geopolitical factors are reshaping where AI labs source compute. For anyone tracking the business infrastructure layer underneath frontier AI development — CoreWeave, Anthropic, Meta, and the capital markets that fund it all — this interview provides unusually direct insight from a decision-maker at the center of these deals.
📺 Source: Bloomberg Technology · Published April 10, 2026
🏷️ Format: Interview
