According to the Financial Times, SpaceX is seeking to raise approximately $10 billion through bank loans and an additional $30 billion through investment-grade debt.

Apollo Global Management is expected to lead this financing, and Pimco is also discussing participating in it.

This transaction is still in its early stages, and if it is ultimately completed, it is expected to be finalized in 2027.

The expansion of AI under SpaceX is becoming increasingly dependent on capital.

This financing will further deepen the transition of SpaceX from rockets and Starlink to large-scale AI infrastructure.

The company has signed several important contracts to sell computing power to external customers. A computing power agreement with Google allows Google to use approximately 110,000 units of Nvidia GPU per month, at a cost of 920 million US dollars.

SpaceX subsequently signed another contract worth approximately $1.11 billion per month, which boosted management's confidence in achieving their annual revenue target of $100 billion. The new AI computing power transaction indicates that the additional GPU production capacity has become strategically important.

This creates a direct financing cycle: SpaceX borrows billions of dollars to purchase Nvidia hardware, deploys this hardware in data centers, and then attempts to recoup the investment through long-term computing power contracts.

AI Borrowings have soared from $20 billion to $88 billion

SpaceX The proposed financing occurs at a time when the debt ratio significantly increases during the AI investment cycle.

According to Reuters, loans related to AI in the U.S. leveraged financing market have risen from approximately $20 billion at the beginning of 2025 to $88 billion in 2026, representing an increase of about 340%.

This figure does not include the financing packages within the SpaceX plan, as those are considered investment-grade financing and not part of the leveraged financing that constitutes the $88 billion statistic. However, both indicate the same broader trend: AI infrastructure increasingly relies on external capital, rather than just relying on the company's own cash.

As previously reported by Coinpaper, infrastructure debt from AI is now spreading to GPU suppliers, data center operators, and other companies that have built capacity in advance to meet anticipated demand.