Oracle Promotes Agentic AI For ERP And Supply Chain

Oracle’s decision to rename CloudWorld as AI World says a lot about where the enterprise software market is heading. The focus has shifted from moving workloads to the cloud to building AI directly into how businesses run. The change from CloudWorld to AI World isn’t just about branding; it’s about Oracle competing on intelligence rather than just infrastructure.

As AI World 2025 unfolds in Las Vegas and online, the company is emphasizing how agentic AI is moving from pilot projects into everyday ERP and supply chain workflows. Inside Oracle Fusion Applications, agents are being designed to take on operational work — processing invoices, managing exceptions, onboarding suppliers, maintaining compliance and so on. It’s a practical shift to make AI an active part of business operations, complete with scale, trust and transparency. In this context, Oracle’s overarching message is that AI belongs where complexity meets execution. (Note: Oracle is an advisory client of my firm, Moor Insights & Strategy.)

Oracle’s product evolution comes in the context of rapid adoption of AI (especially agents) across the ERP sector, with significant product changes also being made by SAP, Microsoft and other competitors. Oracle’s aggressive introduction of AI for ERP also fits into the company’s broader ambitions for growth in AI, as reflected, for example, by its $300 billion contract to supply cloud services for OpenAI.

Oracle AI Agents Push ERP And Finance Into New Functional Territory

This year, Oracle is expanding how embedded AI works inside Fusion Cloud ERP and EPM. For example, the Payables Agent automates the invoice process — intake, matching, tax checks and approvals. The Ledger Agent aims to help accountants shift from report chasing to having continuous insight updated with real-time adjustments, and the Planning Agent supports event-driven forecasting to help finance teams respond faster to change.

Because these agents are built directly into Oracle’s cloud applications, companies won’t have to purchase or integrate extra modules. That makes adoption more straightforward and should enable finance teams to make a measurable impact on operational efficiency without adding complexity.

Hari Sankar, Oracle’s group vice president for EPM, summed it up nicely: “With agentic AI, we’re at a point where technology can touch not just isolated workflows, but fundamentally reinvent how accounting and planning teams synchronize, analyze and act on their data.”

How AI Agents Could Transform Supply Chain Practices

Supply chain is another major focus this year. New agents in Fusion Cloud SCM automate key planning, fulfillment and compliance processes that have previously required manual oversight. The Product 360 Advisor Agent flags recalled inventory and suggests alternate sourcing before shipments are disrupted. The Quote-to-Purchase Requisition Agent converts supplier emails into requisitions automatically. Maintenance and Disposition Agents create and assign work orders, while the Item Shortages Analysis Agent identifies disruptions and recommends substitutes to keep production moving. Together, these agents aim to make supply chain operations more proactive and connected.

This is an area where Oracle’s integration of the Model Context Protocol industry standard released earlier this year should be particularly helpful; MCP’s interoperability across different vendors’ systems enables AI agents to pull live data from supplier, carrier and partner systems through no-code configuration. That’s a valuable capability for companies operating in hybrid or multi-system environments, where interoperability is often the hardest challenge to solve.

Derek Gittoes, Oracle’s vice president for SCM, framed the benefits like this: “Inventory planners, procurement teams and fulfillment managers can now automate supply chain processes that previously took hours — sometimes days — to resolve.” He specifically highlights the role of AI agents. “Agents handle exceptions, summarize shortages and recommend alternatives in real time, connecting planning with execution for faster, smarter decisions.”

The New Oracle AI Agent Marketplace And Ecosystem

The Fusion Applications AI Agent Marketplace — part of the Oracle AI Agent Studio — is one of the big announcements coming out of the event. It allows customers to deploy Oracle-validated or partner-built agents directly within Fusion — no code required, and with the intention of no integration headaches.

Teams using the Marketplace can experiment with different agents for functions such as sustainability tracking, supplier risk scoring or asset maintenance, and compare performance without long development cycles. That flexibility could make AI adoption more practical for large organizations that want results, not prototypes.

All agents available in the Marketplace, including those from partners, are validated for audit and compliance at enterprise scale, Oracle says. With more than 32,000 certified partners supporting the ecosystem, customers will have a broad choice of options for configuration and optimization. If the Marketplace delivers as Oracle expects, it could be a real bridge between the company’s AI vision and real-world execution for customers.

Industry Outcomes And Remaining Challenges

Oracle is also previewing early customer results that show how agentic AI can deliver measurable gains across industries. In manufacturing, companies are using agents to track expiring materials, cut waste and improve compliance. In logistics, agents are automating outbound shipment checks to meet export regulations. In healthcare, providers are gaining stronger audit trails and real-time visibility into inventory to maintain continuity of treatment and meet changing regulatory standards.

That said, the road to adoption isn’t frictionless. Integration with legacy systems, maintaining strong governance and navigating organizational change remain hurdles. Many companies are taking a phased approach, deploying a limited number of agents at first to modernize specific workflows before committing to a complete system overhaul. MCP-enabled agents could play an important role here, connecting Oracle and partner data to deliver early improvements like live shortage alerts and automated sourcing. That could establish a practical, hybrid path toward AI adoption that helps enterprises evolve at their own pace.

Oracle Focuses Its Agentic AI Roadmap On Customer Outcomes

Oracle is focused on AI that’s secure, explainable and auditable without slowing down the pace of work. Every agent, whether built by Oracle or a partner, must meet transparency and accountability standards. The company has also given demonstrations of how agent decisions can be traced and reviewed for compliance, which is especially important for companies in industries like finance and healthcare that face constant oversight.

What’s interesting is how Oracle is positioning itself compared with its peers. Microsoft focuses on productivity with Copilot, SAP has doubled down on governance and Oracle focuses on intelligent and accountable AI. It’s a practical stance intended to meet many companies where they are right now — curious about AI, but still cautious about control and compliance.

The bigger test will come after the event, when it’s time to help customers operationalize agents to genuinely streamline ERP and SCM without creating new layers of complexity. Oracle will need to prove that its embedded approach can scale across industries while keeping governance intact. If it succeeds, we may look back on AI World 2025 as a point when Oracle enterprise systems started operating with intelligence built in, not bolted on.

Forbes