Oracle is overhauling its Fusion finance and procurement software to work more directly with AI agents, shifting the user experience away from forms and menu trees and toward business questions, recommendations, and automated task execution.
The move is aimed at large enterprise customers using Fusion for core operational processes. The idea is that users should be able to ask questions such as how to reduce the cost of a product design, accelerate delivery, or cut supply chain risk, while AI agents gather data, assemble context, and carry out more of the underlying work across Oracle applications and connected third-party systems.
That marks an important change in how enterprise software is being presented to buyers. Instead of treating AI as an assistant that sits on top of existing workflows, Oracle is reworking the interface of ERP so that agents become part of how finance and procurement teams actually use the system. In practice, that means less time spent navigating screens and entering routine data, and more reliance on software to surface options, draft actions, and move tasks through approval paths.
Steve Miranda, executive vice president of applications development at Oracle, told Reuters: “Typing in an invoice isn’t a particularly high-value skill to your enterprise or to the person you know who does that part of their job.” The comment captures Oracle’s wider argument: repetitive execution can move to AI, while people remain responsible for negotiation, oversight, judgment, and risk trade-offs.
The strategy also builds on groundwork Oracle has already laid. Over the past several months, the company has introduced embedded AI agents across Fusion for finance, HR, supply chain, sales, marketing, and service. Its own product materials show how specific those use cases have become. In finance, agents can support invoice intake, matching, routing, journal monitoring, and planning analysis. In procurement, they can compare products, interpret supplier quotes, and help turn those quotations into draft requisitions for review.
Just as important is the governance message. Oracle is not only promising speed and productivity. It is also trying to reassure customers that these agents will operate inside the security, permissions, and approval controls that already govern enterprise systems. That matters because finance and procurement leaders are unlikely to hand execution to AI without a clear audit trail and defined access controls. Oracle’s positioning suggests it believes the strongest defence for incumbent ERP vendors is to make agents native to the record system rather than external to it.
The announcement comes at a time when many investors and software buyers are still working out whether AI will make traditional enterprise applications more valuable by increasing automation, or less valuable by weakening the role of the interface itself. Oracle’s answer is clear: enterprise software remains central if it becomes the layer where agents act. The latest changes to Fusion therefore look less like a standalone feature update and more like an attempt to redraw the operating model of finance and procurement software around AI.
You must be logged in to post a comment.