Oracle moves to raise $50B via debt and equity for massive AI data center expansion amid AI spending scrutiny

Oracle is preparing one of the largest capital raises ever attempted by a software company, lining up as much as $50 billion in debt and equity to keep pace with demand for its cloud infrastructure. The plan, disclosed Sunday, lays bare how quickly AI workloads are reshaping balance sheets across Big Tech.

The company says the funds will be used to add capacity across Oracle Cloud Infrastructure, where demand has surged as customers build and train large AI systems. Those customers include Nvidia, Meta, OpenAI, AMD, TikTok, and xAI.

“Oracle is raising money in order to build additional capacity to meet the contracted demand from our largest Oracle Cloud Infrastructure customers, including AMD, Meta, NVIDIA, OpenAI, TikTok, xAI, and others,” the company said in a statement.

The scale of the raise stands out. The timing matters just as much.

How Oracle Plans to Fund the Expansion

Oracle expects to raise roughly half of the capital through equity-linked instruments and common stock, including mandatory convertible preferred securities and a new at-the-market equity program that could raise up to $20 billion. The rest would come from senior unsecured bonds issued early next year.

Chairman Larry Ellison has long argued that Oracle’s vertically integrated cloud model gives it an edge over hyperscalers that rent capacity from third parties. This financing push suggests that advantage comes with increasing capital intensity, especially once AI training clusters are introduced.

Shares of Oracle fell about 4% in premarket trading after the announcement, a sign that investors initially welcomed clarity on how Oracle plans to pay for its ambitions, Reuters reported.

Investor Anxiety Is Growing Alongside AI Spending

This capital plan lands at an uneasy moment. Two months ago, Blue Owl Capital stepped away from talks to finance a planned $10 billion Oracle-backed data center in Michigan. That pullback followed growing unease across markets about the pace of AI infrastructure spending and the timeline for returns.

Oracle’s debt load has climbed, and its exposure to OpenAI continues to draw attention. OpenAI remains unprofitable and has shared few details about how it expects to fund the compute required for future models. That link has turned Oracle’s cloud strategy into a proxy bet on the economics of frontier AI.

Bond investors have taken notice. In January, Oracle faced a lawsuit from bondholders who claimed the company failed to disclose how much additional debt would be required to build its AI infrastructure. Around the same period, the cost of insuring Oracle’s debt against default jumped to its highest level in at least five years.

A Broader Reckoning for AI Infrastructure

Oracle’s move fits into a wider rush among hyperscalers to secure land, power, and hardware for AI workloads. Data center deals hit a record $61 billion last year, driven by commitments from tech giants racing to avoid capacity shortages.

This funding plan shows how quickly AI has shifted from a product story to an infrastructure arms race. Cloud providers are no longer debating whether demand will arrive. The debate now centers on who can afford to build fast enough without stretching balance sheets too far.

Oracle’s answer is clear. It is willing to raise tens of billions to stay in the fight, even as investors sharpen their focus on risk, leverage, and the true cost of AI at scale.