SoftBank lines up another $30 billion in debt for OpenAI after selling off assets

SoftBank is tying its financial strategy ever more tightly to OpenAI. The Japanese conglomerate has already sold other assets to take part in OpenAI's funding rounds. Now it is preparing to borrow another $30 billion and to revise its own debt rules to do so.

A new round of borrowing

The situation shows how expensive the generative AI bet has become, even for a player of SoftBank's size. Late last year, the company had to hastily sell off some of its other assets to close OpenAI's previous funding round. Now it is looking at fresh borrowing of up to $30 billion. This doesn't look like ordinary investing out of free cash flow. The company is clearly hunting for capital to back one specific strategic bet, and it is willing to rebuild its financial structure quickly to get access to the next stage of the deal.

That sends an important signal to the market. When a major investor isn't simply putting surplus capital into a promising asset, but has to free up liquidity and take on more debt, the perception of risk changes. This is no longer about "adding one more AI investment to the portfolio." It's about concentrating on one of the most expensive areas of the tech market. SoftBank is effectively showing that it ranks its stake in OpenAI so highly that it is willing to sacrifice flexibility elsewhere in the business.

The cost of the AI bet

The core problem with this strategy is that soaring interest in AI doesn't change the basic math of a balance sheet. Every additional loan increases the load on the company: servicing the debt gets more expensive, room for new moves shrinks, and sensitivity to the market grows. If a single deal forces you to sell assets today and change internal limits tomorrow, the project matters not only as a financial investment but also as a symbolic bet on the future of the entire group.

• a rushed sale of some assets to fund OpenAI's previous round
• preparation of new debt of up to $30 billion
• a revision of internal borrowing standards
• an even tighter link between SoftBank's strategy and the success of the AI market

It's the combination of these factors that makes the news stand out. Large-scale debt on its own is no sensation for a global conglomerate. But when money is raised for the same AI asset repeatedly and on an accelerated timeline, investors start looking not only at potential returns but also at the cost of getting it wrong. If the AI market keeps growing, the bet may pay off. If growth slows or valuations start to cool, the debt load will become a far more painful issue.

What changes internally

The plan to revise internal standards for borrowed funds is telling in its own right. Rules like these usually exist precisely to keep a company from concentrating too much risk during periods of market euphoria. If they are being adapted for a new investment, the old framework is already getting in the way of closing the deal at the desired size. That matters more than the debt itself: SoftBank isn't just changing its source of funding, it is changing the internal discipline that determines how much risk the group is willing to take onto its balance sheet.

For OpenAI, by contrast, this is a strong positive signal: it still has an investor willing to find money even under tight constraints and to back its interest with billions in commitments rather than words. For SoftBank, the picture is more complicated. The more the conglomerate ties its capital, debt, and governance rules to a single AI asset, the less room it has to maneuver if market conditions change. This setup could deliver outsized returns, but the price of entry already looks very high.

What it means

The race for stakes in the leading AI companies is entering a phase where even the largest investors have to pay not only with money but with the stability of their balance sheets. SoftBank's story shows that taking part in the AI boom no longer looks like an option; it is an expensive strategic bet with real debt consequences.