While Oracle previously outlined plans to cap additional AI-related financing at USD 20 billion through fiscal 2030, the company is now moving forward with an immediate USD 20 billion fundraising campaign through bond and share offerings to finance its AI and cloud infrastructure expansion. This fresh capital push follows an earlier USD 20 billion share sale completed between June and August 2026, intensifying the pace of its balance sheet expansion and forcing a re-evaluation of how legacy enterprise tech vendors fund the physical hardware layer of the generative AI boom.
According to regulatory filings reported by Reuters, Oracle plans to raise USD 10 billion through a bond sale and another USD 10 billion through a share offering. The two transactions are separate, meaning either offering can proceed independently based on market demand and institutional appetite. This aggressive financing push underpins the physical buildout required to support compute-heavy enterprise workloads, as Oracle builds out facilities alongside AI partnerships like AstraZeneca and Oracle joining hands to use AI in clinical trials.
Oracle plans to raise USD 10 billion through a bond sale and another USD 10 billion through a share offering.
Large-scale data centre commitments demand continuous liquidity, forcing legacy infrastructure providers to utilize both equity and debt markets simultaneously to avoid choking operational cash flow. Oracle has now begun the issuance of USD 10 billion worth of investment-grade bonds, which have been split into six different tranches with maturities spread between 2029 and 2046. This multi-tranche structure spreads out the redemption schedule, but the company has not as yet provided any information on the exact interest rate or amount being offered for each tranche, leaving the precise borrowing cost undisclosed while debt servicing obligations scale up significantly over the next two decades.
OpenAI Commitments and Balance Sheet Strain
Can Oracle sustain its escalating leverage while underwriting physical capacity for major generative model developers without locking in immediate cash returns?
As per the Oracle earnings released on August 31, the debt level stood at about USD 125.34 billion. Oracle is investing heavily in data centres to support major AI customers, including OpenAI, directly linking its balance sheet exposure to the computational scale demanded by foundational model creators. Tech enterprises are turning to debt and equity markets to finance expensive AI infrastructure projects, absorbing massive fixed obligations before generating operating returns from the underlying hardware leases.
The debt level stood at about USD 125.34 billion, with capital expenditure peaking across two fiscal years.
The thesis supporting such leverage assumes that massive capital injections and contracts with high-profile clients like OpenAI justify the debt load, positioning the provider to capture long-term cloud revenue. Yet the financial timeline reveals a prolonged lag between capital deployment and cash generation, stretching the limits of traditional enterprise finance. Oracle anticipates its capital expenditure will hit a high point in the current fiscal year and the next one, creating a multi-year window of peak spending before free cash flow turns positive in the second half of fiscal 2028.
Oracle has committed to multi-decade debt maturities extending out to 2046 to build facilities for today's AI workloads, while operating cash flow to cover those liabilities remains deferred until the second half of fiscal 2028. This structural mismatch between immediate capital expenditure and delayed cash generation highlights the financial risk undertaken by infrastructure providers racing to service the AI market.
What’s behind it
Bond tranches maturing as late as 2046 clearly demonstrate the real price of staying relevant in the data center race. Oracle is essentially pushing balance-sheet risk decades into the future just to deploy immediate compute for clients like OpenAI and AstraZeneca. By late September Big Tech was projected to pour $1.2 trillion into AI infrastructure by 2027, and Oracle now faces the friction of managing roughly $125.34 billion in debt while cash flow recovery remains years away. While frontier model developers burn through rented capacity searching for viable unit economics, the enterprise vendor has already locked in debt obligations spanning the next two decades.
Our stories: Big Tech $1.2T AI Spending Plan Outpaces Revenue Growth (September 27).