ORACLE Q4 FY26 earnings call ORCL

Oracle is uniquely positioned to help customers leverage AI by delivering applications, data, infrastructure, AI tooling, and industry expertise together. Customers are moving beyond AI experimentation to implement enterprise-grade, agentic solutions to run their businesses. Over the past year, Oracle has delivered more than 1,000 AI agents across its application suites, which can reason, decide, and execute work across processes. This integration allows customers to quickly and affordably consume AI within their existing Oracle applications.

A key focus for customers is leveraging their own proprietary data sets with AI, much of which already resides in Oracle databases or is generated by Oracle applications. Inferencing against decades of rich operations data can exponentially compound the benefits of AI. Oracle's full-stack offerings enable customers to quickly get up and running, combining AI with their private data sets, which is a critical differentiator in the market.

Oracle Cloud Infrastructure (OCI) is designed for high security, performance, flexibility, and low cost, achieved through continuous innovation across all layers. OCI has been the fastest-growing cloud provider, and the AI infrastructure market is projected to be trillions of dollars annually, making the existing cloud market seem small by comparison. Oracle believes OCI will grow into an extremely large and profitable business with a 30% to 40% margin profile.

Oracle is making significant capital investments to capitalize on growth opportunities, with a net cash outlay for capital expenditures of $48 billion in FY26 and an expected $70 billion in FY27. These investments are driven by committed customer demand, reflected in record RPO, and are expected to yield strong returns on invested capital in the high 20s at a steady state for infrastructure projects. Customer prepayments and bring-your-own-hardware models contribute to favorable margins and funding requirements.

Oracle is simplifying AI consumption and payment through new agentic pricing models. While much AI innovation in core applications is included at no extra charge, customers can purchase additional agentic capacity via token bundles for advanced reasoning. Outcome-based commercial models are also being introduced, aligning pricing directly to value derived, such as interview agents priced by candidates screened or hospitality upsell agents by transaction percentage. This approach helps customers control costs and align spending with generated value.

Oracle is making massive progress on its data center build-out, delivering over 1.2 gigawatts of capacity in FY26, with FY27 Q1 expected to approach 1 gigawatt. Key sites like Abilene, Texas, have delivered 42% of total capacity, with an additional 35% expected in the next 90 days. Other sites like Shackelford, Texas, and Doña Ana County, New Mexico, are on track for customer delivery in H1 CY27, with significant power capacity already online or designed with clean energy solutions.