OpenAI's Planned Cloud Spending Balloons to $750 Billion Through 2030

OpenAI has raised its projected cloud computing spending to $750 billion through 2030, a dramatic increase from the $600 billion figure previously reported, according to an exclusive Wall Street Journal report published July 22, 2026.

The revised figure highlights the staggering scale of the AI infrastructure arms race and raises fresh questions about the financial sustainability of the largest AI companies. Stay current with our latest AI developments for ongoing coverage of this story.

The Numbers Behind the Spending

According to the WSJ report, which was corroborated by Yahoo Finance, TechCrunch, and Quartz, OpenAI's compute spending plans now total approximately $750 billion over the remainder of the decade. TradingKey reported that the company also plans to spend $20 billion building its own data centers as part of this commitment.

The previous projection of $600 billion had already stunned investors and industry analysts. The $150 billion increase reflects the accelerating pace of AI model development and the company's ambition to build proprietary infrastructure rather than relying entirely on cloud providers.

Moomoo reported that the $750 billion figure represents OpenAI's cumulative cloud spending commitment through 2030, encompassing agreements with Microsoft Azure, CoreWeave, Oracle, and other infrastructure providers, as well as the company's own data center construction.

Why the Costs Keep Climbing

The soaring spending projections reflect several converging factors in the AI industry. Frontier AI models — the most advanced systems from OpenAI, Google, Anthropic, and others — require enormous computational resources for both training and inference. Each new model generation typically demands an order of magnitude more compute than its predecessor.

OpenAI's GPT-5 and GPT-5.5 model families, launched in 2026, have pushed the boundaries of model size and capability. The company's agentic AI products, which allow models to autonomously execute multi-step tasks, require significantly more inference compute than traditional chatbot interactions. As OpenAI expands into autonomous agents, coding assistants, and enterprise applications, the demand for inference capacity grows in lockstep.

The company has also been investing in physical infrastructure. Building its own data centers — at an estimated $20 billion cost — represents a strategic shift away from pure reliance on Microsoft Azure. This vertical integration could reduce long-term costs but requires enormous upfront capital.

Financial Sustainability Questions

The $750 billion spending plan has reignited debate about the financial viability of frontier AI companies. Leaked audited financial statements revealed that OpenAI's operating losses swelled to $20.9 billion in 2025, even as revenue tripled to $13 billion. The gap between revenue and spending has widened dramatically as compute costs escalate.

OpenAI's CFO Sarah Friar has publicly pitched a "useful-intelligence-per-dollar" framework for measuring AI returns, arguing that companies should evaluate AI spending by work accomplished rather than per-seat licensing. This framing suggests OpenAI is working to justify its massive capital expenditures to investors and enterprise customers.

Critics, including some prominent economists and AI researchers, have warned of an emerging "AI debt tsunami" — a reference to the hundreds of billions in capital being poured into AI infrastructure with uncertain returns. Morgan Stanley and Goldman Sachs have separately flagged the risk of overinvestment in AI data centers if demand for AI services fails to meet projections.

The Competitive Landscape

OpenAI is not alone in its massive spending commitments. Anthropic recently announced a deal to deploy 2 gigawatts of AMD Instinct MI450 GPUs and is in talks to lease compute from Meta in a deal worth approximately $10 billion. Google continues to invest heavily in its custom TPU accelerators and Tensor Processing Units. Meta is spending tens of billions on its own AI infrastructure.

The aggregate spending across the industry now exceeds $1 trillion when combining the commitments of OpenAI, Google, Anthropic, Meta, Amazon, and Microsoft. This level of investment has drawn comparisons to the telecommunications buildout of the late 1990s, which was eventually justified by internet adoption but bankrupted many participants along the way.

Infrastructure Buildout Accelerating

OpenAI's spending is spread across multiple infrastructure projects. The company has secured cloud capacity from Microsoft Azure, CoreWeave, and Oracle, and is now building its own data centers to reduce dependency on third-party providers. The $20 billion data center investment, reported by TradingKey, signals a strategic pivot toward owning rather than leasing computing infrastructure.

This shift could give OpenAI more control over its deployment environment and reduce the risk of supply constraints. However, it also increases the company's capital intensity and operational complexity at a time when it is reportedly preparing for a public offering.

What This Means for the Industry

The $750 billion figure, if realized, would represent one of the largest technology infrastructure investments in history. It exceeds the combined cost of the Apollo program, the Interstate Highway System, and the rollout of 4G cellular networks in inflation-adjusted terms.

For cloud providers, OpenAI's push to build its own data centers could reduce future revenue growth from AI workloads. For chipmakers like Nvidia and AMD, the spending commitment signals sustained demand for accelerators through the end of the decade. And for startups and smaller AI companies, the massive capital requirements raise barriers to entry in the frontier model market.

Stay Ahead of AI

The infrastructure spending war is reshaping the entire tech industry. Read more AI news →