Oracle Corp. told investors that surging demand for artificial intelligence infrastructure will sustain its data center expansion through at least 2027, a forecast that sent shares up roughly 8% after the company reported fiscal third-quarter results. The prediction rests on a contract backlog that has ballooned to record levels, giving Oracle unusual visibility into future revenue at a time when rivals are still scrambling to lock in long-term cloud commitments. For investors and enterprise buyers alike, the numbers signal that the AI spending cycle is far from peaking.
A $552.6 Billion Backlog Anchors the Forecast
The single most telling figure in Oracle’s latest quarterly filing is its remaining performance obligations, or RPO, which stood at $55.26 billion as of February 28, 2026. RPO represents contracted revenue that has not yet been recognized on the income statement. In practical terms, it is a measure of how much future business Oracle has already locked in through binding agreements with cloud and AI customers.
That total represents a 325% year-over-year increase, a growth rate that dwarfs anything Oracle posted during prior cloud transitions. The sheer scale of the backlog matters because it converts what would otherwise be a speculative growth story into a contractually grounded one. Customers have signed deals; the question is when, not whether, that revenue arrives.
Oracle’s SEC filing breaks down the recognition timeline: 12% of the total is expected within the next 12 months, 31% over months 13 through 36, and the remainder stretching out beyond three years. That schedule means roughly $6.6 billion should convert to revenue within a year, with another $17.1 billion arriving in the two years after that. The long tail of recognition, with more than half of the backlog tied to periods beyond 36 months, suggests Oracle’s AI data center contracts are structured as multi-year commitments rather than short-burst deals.
Revenue Outlook Climbs to $90 Billion
Alongside the backlog disclosure, Oracle raised its revenue forecast, projecting $90 billion in annual sales by fiscal year 2027. That target reflects a bet that AI workloads will continue to expand faster than Oracle can build capacity to serve them. The company framed the gap between demand and available supply as a structural tailwind rather than a temporary spike.
The distinction matters. Many technology companies have posted strong AI-related quarters only to see growth moderate as initial deployment waves cool. Oracle is making a different argument: that its contract structure, heavy on long-term infrastructure commitments from hyperscale customers, insulates it from the short-cycle volatility that has historically whipsawed enterprise software stocks. Whether that insulation holds depends on whether the companies signing these deals continue to need the capacity they have reserved, a risk that grows if AI adoption plateaus or if cheaper alternatives emerge.
What RPO Actually Measures, and What It Does Not
Remaining performance obligations are not the same as guaranteed revenue. As Oracle disclosed in its fiscal year 2024 annual report, the figure can shift due to foreign-exchange fluctuations, changes in contract mix, and seasonal patterns in enterprise buying. Contracts can also be renegotiated or, in some cases, terminated, though penalties for early exit typically discourage cancellation.
The gap between RPO and actual revenue recognition is worth watching closely. A company can report enormous backlog growth while still facing quarters where revenue comes in below expectations if contract ramp-ups are delayed or if customers consume capacity more slowly than projected. Oracle’s 12% near-term recognition rate, for instance, means that only a fraction of the $552.6 billion will show up on the income statement in the current fiscal year. Investors treating RPO as a simple revenue proxy risk overestimating the pace of cash conversion.
Shares React to the AI Demand Signal
Markets responded quickly. Oracle shares rose about 8% following the earnings release, and were up about 10% in premarket trading by Wednesday morning, according to The Wall Street Journal. The rally reflected relief as much as enthusiasm. Heading into the report, investors had debated whether AI infrastructure spending was peaking.
Oracle’s results cut against that narrative. The 325% RPO surge and the raised revenue target gave investors evidence that at least one major infrastructure provider is seeing demand accelerate, not flatten. The stock move also reflects Oracle’s evolving competitive position. Once viewed primarily as a database and enterprise applications company, Oracle has spent the past several years building out cloud regions to compete more directly with Amazon Web Services and Microsoft Azure for large-scale workloads.
Competitive Pressure and Consolidation Risk
The scale of Oracle’s backlog raises a broader question about the AI infrastructure market: can smaller providers keep up? When a single company reports more than half a trillion dollars in contracted future revenue, the implication is that enterprise buyers are concentrating their spending with a handful of vendors capable of delivering capacity at global scale. That dynamic could accelerate consolidation among mid-tier cloud and data center operators that lack the balance sheet to match Oracle’s buildout pace.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
