Sources: SoftBank plans to borrow $16B to invest in AI, and might borrow another $8B in early 2026, which could strain its already debt-heavy balance sheet
The biggest risk-taker in tech investing is back to his old playbook, loading up on debt to fund his latest obsession: artificial intelligence.
Context & Ripple Effects
SoftBank’s AI spending had already more than doubled to $8.9B over the preceding 12 months, marking a shift from a stated investment “counteroffensive” toward a capital-intensive AI push. Reports of prospective OpenAI and Stargate commitments made funding capacity central to that strategy.
The proposed borrowing puts SoftBank’s discounted portfolio value and debt load at the center of its ability to keep financing AI investments. It matters because the company is pairing large AI ambitions with balance-sheet leverage rather than relying solely on asset sales or operating cash flow.
First-order effects
- SoftBank would gain financing capacity for AI investments, while increasing interest, refinancing, and balance-sheet pressure immediately.
- Lenders and SoftBank shareholders would have greater exposure to the value of the company’s holdings, particularly if further borrowing is required.
Second-order effects
- A debt-funded SoftBank could sustain competition for major AI investments and infrastructure commitments, raising the importance of financing terms alongside investment selection.
- The move reinforces demand for large bridge and asset-backed financings; SoftBank subsequently sought a bridge loan of up to $16.5B for US AI investments, illustrating how funding execution can become a separate constraint.
Third-order effects
- If repeated, this model would make access to collateral, credit capacity, and structured finance a more important divider between AI investors than cash on hand alone.
- It also concentrates AI-investment risk in a smaller set of highly leveraged capital providers, making portfolio-value swings more consequential for the pace of AI deployment.
The trend: AI investment is evolving into a compute-finance cycle in which debt capacity and asset collateral increasingly determine who can fund large-scale AI bets.