From Hype to Hard Work — Oracle AI World 2025 and the Future of Intelligent Enterprise.

Every keynote, every breakout, every customer panel orbited the same message: AI is not something you experiment with on the side anymore. It’s how core systems run, how users work, how data is governed, and how infrastructure is designed, powered, and priced.

Oracle AI World 2025 was the company’s first major global event under the co-leadership of Mike Sicilia and Clay Magouyrk. Together with Chairman and CTO Larry Ellison and Applications EVP Steve Miranda, the new leadership team delivered a clear message — AI is now infrastructure.

Mike Sicilia: AI as an Enabler, Not a Replacement

Mike Sicilia, Oracle’s EVP of Global Industries (and now co-CEO), opened with what felt like a correction to the entire market narrative.

“AI is an enabler,” he said. “It’s helping people do their work better, not replacing them.”

Instead of product slides, Sicilia’s keynote featured real-world customer voices. These weren’t pilots or proof-of-concepts — they were enterprise AI deployments already transforming daily operations across critical sectors.

Customer Spotlights

• Exelon – Predictive grid analytics powered by Oracle AI models analyzing weather, equipment, and usage data to forecast potential outages before they happen. By integrating AI with Oracle Utilities and OCI Data Science, Exelon is increasing reliability for over ten million customers.
• Avis Budget Group – Using Oracle Database 23AI, employees can query data conversationally without writing SQL. Queries like “Which vehicle classes are generating the highest margin this quarter?” return instant, context-aware results. CIO Ravi Simhambhatla coined a new definition: “AI doesn’t mean Artificial Intelligence; it means Augmenting Individuals.”
• Marriott International – Leveraging AI with Oracle Fusion Applications, Marriott unified HR, finance, and property systems across hundreds of hotels. The result is smoother data flow, reduced manual entry, and more time for staff to deliver personalized guest experiences — proving AI’s human-centric purpose.
• Biofy (Brazil) – Using Oracle’s vector database and AI-powered search, the biotech firm cut antimicrobial resistance diagnosis from five days to just four hours, supporting life-saving healthcare interventions.

These industries — utilities, transportation, hospitality, and life sciences — are all heavily regulated, high-risk environments. Oracle’s deliberate choice of customers showcased one point clearly: AI is enterprise-ready, compliant, and secure at scale.

Larry Ellison: AI After the Hype Cycle

Larry Ellison’s two-hour keynote was part technical deep-dive, part declaration. He drew a clear line between the first era of AI — model training on public internet data — and the second era: secure reasoning on private enterprise data.

“People want models to reason on their private data, but still keep it private,” Ellison said. “They want to have their cake and eat it too.”

The Oracle AI Data Platform

Oracle’s solution to that paradox is the Oracle AI Data Platform, combining the AI Database, retrieval-augmented generation (RAG), vector search, and fine-grained policy control. Together, these enable models to “think with” enterprise data without copying it or retraining on it.

That architecture directly addresses enterprise AI’s biggest blockers:

1. Data Residency – Sensitive data remains within required geographic or legal boundaries.
2. Audit & Access Control – All AI activity is logged and explainable.
3. IP Protection – Proprietary data isn’t exposed to public models.

In Ellison’s words: “The model can reason across your data, not look at it.”

Oracle Builds Oracle With AI

Ellison also revealed that Oracle engineers are now building software this way internally. Using Oracle APEX and a declarative AI generation model, developers define intent and constraints while AI generates secure, stateless, production-ready code.

This approach has already delivered massive results: Oracle modernized Cerner’s 25-year-old healthcare system in just three years. The proof point is profound — Oracle is not just selling AI tools, it is using AI to build Oracle itself.

Clay Magouyrk: AI Runs on Physics, Not Slogans

Day 2 shifted gears from AI’s purpose to its plumbing. Clay Magouyrk, EVP of Oracle Cloud Infrastructure (OCI) and co-CEO, emphasized that none of these ambitions matter without scalable compute, network speed, and energy availability.

Acceleron and Superclusters

Magouyrk unveiled Acceleron, OCI’s next-generation input/output architecture — a multi-plane, low-latency fabric that eliminates traditional network bottlenecks. By removing “middle boxes” and streamlining data flow, Acceleron allows for 100G and 400G connections capable of supporting massive AI workloads.

Oracle also announced its Zettascale10 supercluster, delivering up to 16 zettaFLOPS of compute with 800,000 NVIDIA GPUs and AMD Instinct accelerators. This scale supports continuous AI training and inference cycles at a lower total cost of ownership.

The World’s Largest AI Data Center

Oracle’s partnership with OpenAI extends beyond compute provisioning. Together, they’re building a 1.2-billion-watt campus in Abilene, Texas, the largest AI data center on the planet. That’s enough to power industrial-scale LLM reasoning while maintaining energy efficiency through OCI’s renewable energy strategy.

These engineering feats aren’t theoretical. TikTok’s Head of Infrastructure joined Magouyrk on stage, describing the collaboration with Oracle as “a joint stability mission.” OpenAI’s Peter Hoeschele also highlighted Oracle’s co-engineering role, not just as a capacity provider but as an innovation partner.

Multicloud Universal Credits

Oracle doubled down on its multicloud philosophy, announcing Multicloud Universal Credits — a single pool of credits usable across OCI, Azure, Google Cloud, and AWS. This simplifies procurement while ensuring customers can run workloads wherever they need, meeting regional compliance requirements without vendor lock-in.

It’s both a technical and financial innovation — the kind of flexibility enterprise buyers have been demanding for years.

Steve Miranda: AI Inside the Apps You Already Own

If Ellison focused on AI’s why and Magouyrk on its how, Steve Miranda focused on its where: directly inside the software people use every day.

Miranda demonstrated embedded AI agents within Oracle Fusion Applications, showing how they automate repetitive work and surface actionable insights across finance, HR, supply chain, and operations.

AI Agent Marketplace and Embedded Agents

Oracle has now delivered 400+ AI agents, all pre-trained with Oracle’s domain models and integrated security. Additionally, the new Fusion AI Agent Marketplace enables partners to publish certified, industry-specific agents for niche business processes.

These agents aren’t experimental. They reconcile ledgers, flag anomalies, answer employee policy questions, summarize supplier risks, and even simulate financial scenarios — all within Fusion. And the biggest announcement? They come at no additional license cost.

This pricing move redefines how enterprises value AI: it’s no longer a premium feature; it’s a baseline entitlement. As Miranda said, “You shouldn’t have to move your data to get value from AI. AI should come to where your data already lives.”

Where This Leaves Customers and Partners

Oracle AI World 2025 marked a genuine turning point: AI has moved from pilot to platform. Oracle’s integrated ecosystem — from infrastructure to data to applications — gives enterprises a secure, governed, and scalable foundation to operationalize AI.

For CIOs, this means faster ROI, predictable cost, and built-in compliance. For partners, it opens new opportunities to build industry-specific agents, automate cross-system workflows, and deliver measurable outcomes.

The broader message was unmistakable: the AI hype cycle is over. The implementation era has begun. Oracle is betting on integration, trust, and real enterprise scale — and so far, it’s a bet that’s paying off.

If you didn’t get a chance to connect with our team at AI World — or you’re now planning your 2026 AI roadmap — schedule a meeting with us. We’ll help you map Oracle’s AI stack to your business priorities, identify where to start delivering value, and make AI work safely, efficiently, and at scale.