Masayoshi Son is betting heavily on artificial intelligence — and this time, bond investors are being asked to help pay the bill.
SoftBank Group has launched what is set to be one of the largest corporate high-yield bond offerings ever, seeking more than $11 billion from debt investors as the Japanese conglomerate accelerates its financing push around artificial intelligence.
The offering consists of approximately $10 billion in U.S.-dollar bonds and €1 billion, or roughly $1.1 billion, in euro-denominated debt, according to Bloomberg. The deal is designed to provide financing for SoftBank's enormous commitments to the AI sector, including its investment in OpenAI.
The scale of the transaction tells a bigger story than SoftBank alone.
AI has become one of the largest capital-allocation themes in global financial markets, and companies are increasingly turning to debt markets to finance the infrastructure, acquisitions and strategic investments required to participate in the boom.
SoftBank is now one of the clearest examples of that trend.
The Japanese investment giant has already committed tens of billions of dollars to artificial intelligence. Its commitments to OpenAI are approaching $65 billion, while the company is also exploring additional mergers, acquisitions and investments across the sector.
Reuters reported that the new bond offering includes a $10 billion portion in U.S. dollars and €1 billion in euros. The dollar notes are being offered across 3.5-year, 5.5-year and 7.5-year maturities, while the euro portion includes four-year and six-year maturities. The proceeds are expected to help finance SoftBank's $10 billion payment for the third tranche of its OpenAI investment, with the deal expected to close on October 1.
There is another reason this bond sale matters.
It demonstrates how the AI investment boom is increasingly moving from equity markets into credit markets.
Investors have spent years buying shares of chipmakers, cloud companies and AI software businesses in anticipation of rapid growth. Now bond investors are becoming equally important because the companies building and financing AI infrastructure need enormous amounts of capital.
Goldman Sachs credit strategists estimate that global AI-related debt issuance has already exceeded $575 billion in 2026.
That is an extraordinary amount of borrowing connected directly or indirectly to the technology boom.
The change in financing strategy is understandable.
AI infrastructure is expensive.
Data centers require buildings, power systems, cooling equipment, networking hardware and advanced processors. Developing large AI models requires huge computing resources, while companies competing for market share are spending heavily on talent and acquisitions.
Equity financing can provide capital, but borrowing allows companies and investment firms to fund large projects without immediately issuing additional shares.
SoftBank is taking that strategy to an extreme scale.
The company is effectively asking investors to lend it money based on confidence in the future economic value of its AI investments.
That is where the story becomes more complicated.
SoftBank's executives remain strongly bullish on the sector. Founder and CEO Masayoshi Son has previously argued that AI-related industries could eventually represent around 20% of global economic output by 2040, equivalent to roughly $46 trillion.
Those are enormous expectations.
The financial markets, however, ultimately need those expectations to translate into cash flows.
Bondholders are not simply betting that AI will become transformative. They are lending money and expecting the borrower to continue meeting interest and principal obligations.
That makes debt financing fundamentally different from an equity investment.
An equity investor can theoretically wait years for an AI company to grow into its valuation. A bond investor has a contractual claim that comes due on a defined schedule.
The distinction becomes particularly important as borrowing across the AI ecosystem increases.
If AI revenues grow rapidly and infrastructure investments produce strong returns, the debt can help companies scale more quickly.
If returns disappoint, the same debt can magnify financial pressure.
That is the concern increasingly circulating through credit markets.
Bloomberg reported that the rapid buildup of AI-related debt has made some bond investors uneasy because a growing pile of obligations could create a financial hangover if the technology fails to deliver sufficient returns.
SoftBank itself has taken steps to manage the immediate financing requirements around OpenAI.
Reuters said the new notes are intended in part to replace a $10 billion bridge loan that SoftBank had secured for the investment. Fitch Ratings assigned a BB+ rating to the proposed notes and said the company was expected to maintain adequate liquidity and access to capital markets, while also acknowledging that SoftBank's debt burden would rise.
That rating places the proposed debt in the high-yield category, highlighting the trade-off between the enormous growth opportunity associated with SoftBank's investments and the financial risks associated with leverage.
The size of the offering also gives investors another way to gauge demand for AI-related credit.
For much of the past several years, enthusiasm for AI has been most visible in the stock market. Semiconductor companies have reached enormous valuations, technology giants have expanded capital spending and private AI companies have attracted huge funding rounds.
Now, debt markets are becoming an equally important part of the story.
The question is whether lenders remain comfortable providing money on such a large scale.
So far, the numbers suggest considerable appetite.
The broader AI-related debt boom indicates that investors are willing to finance the infrastructure necessary to build the industry's next stage. But the continued willingness of credit markets to lend will depend increasingly on cash generation, asset quality and the ability of borrowers to service their obligations.
SoftBank's transaction therefore represents both confidence and risk.
The confidence is obvious.
Few companies are willing to raise more than $11 billion in a single high-yield offering unless management believes its long-term investments justify the financing.
The risk is equally clear.
A large amount of debt creates a fixed financial obligation, regardless of whether AI revenues arrive on schedule.
The issue becomes even more important because SoftBank is not making a single AI bet. It has exposure to a broad ecosystem through investments, acquisitions and strategic positions.
If the AI industry continues expanding rapidly, that strategy could provide enormous upside.
If AI spending eventually slows, however, the company could face a more complicated environment in which large investments compete for capital with debt-service requirements.
There is also a broader lesson for the financial system.
Technology booms often start as stories about innovation and eventually become stories about financing.
The internet boom required fiber networks and data centers. The smartphone era required enormous semiconductor and manufacturing investment. AI is following a similar pattern, but the scale is potentially much larger.
The difference this time is that the financing machine itself has become part of the investment thesis.
Companies need capital to build AI infrastructure.
Investors provide that capital because they expect AI to generate future economic returns.
Those returns, in turn, must become large enough to justify the borrowing.
That creates a feedback loop.
As AI spending grows, financing grows with it. As financing grows, companies can spend more. The resulting infrastructure may accelerate AI adoption — but it also increases the financial consequences if expectations prove too optimistic.
SoftBank's jumbo bond sale captures that tension perfectly.
It is simultaneously a vote of confidence in artificial intelligence and a reminder that even the most ambitious technology revolution eventually has to pass through the ordinary mechanics of debt markets.
The AI race may be powered by chips and algorithms.
But increasingly, it is being financed by bonds.