Oracle Launches AI Agents for Enterprise Apps, But Liability Questions Linger

Oracle has unveiled its Fusion Agentic Applications, a suite of AI agents designed to make and execute autonomous decisions within enterprise business processes. The announcement, made in London this week, represents Oracle's latest push into the rapidly evolving AI agent market, where major tech vendors are racing to capitalize on growing enterprise demand for automation.

The new applications will be integrated with Oracle's existing Fusion Cloud Applications suite, covering financials, ERP, HR, payroll, and supply chain management. Oracle's pitch centers on a key advantage: the data needed to train and run these agents already resides within its enterprise applications ecosystem.

Steve Miranda, Oracle's executive vice president of applications development, framed this as a fundamental shift from process-focused software toward outcome-driven automation. "Applications that can reason, decide, and act in pursuit of defined business objectives," Miranda said, describing the new paradigm.

One example Oracle highlighted is the Design-to-Source Workspace Agentic Application, which promises to coordinate decisions across engineering, supplier, and sourcing functions to create a "coordinated and continuous process."

However, analysts are urging caution about the technology's current limitations. Balaji Abbabatulla, Gartner vice president and vendor lead for Oracle, acknowledged that while the concept sounds promising, significant challenges remain under the hood.

"Our position is that this sounds good, but be cautious. It doesn't necessarily look as glittery as it sounds. There are challenges under the hood which are not being overcome right now, but maybe over time," Abbabatulla said.

A major hurdle is data integration. Oracle has launched an AI Data Platform to integrate data from different sources to build AI agents, with the goal of connecting non-Oracle repositories, legacy applications, and extracting information from them. However, Abbabatulla noted that this process isn't automated.

"There's no kind of autonomous way of synchronizing these different data repositories in the background," he explained. Building agents to run application-based processes will require significant work and likely substantial investment in Oracle engineering expertise.

This presents a particular challenge for large enterprises already invested in data platforms from vendors like Databricks, Snowflake, or Cloudera. These organizations have made significant "big data" era investments that Oracle's pitch must overcome.

Abbabatulla sees Oracle's strategy as partly defensive, using data-in-context as an incentive to keep customers within its ecosystem. "The transition overhead is massive, because these are investments people have made for years now," he said. "This is unlikely to actually attract them to let go of this investment, but I'm sure there'll be organizations willing to try this in addition to some of those other investments they have made."

The most pressing unanswered question remains liability. If an AI agent makes a bad decision at scale and speed, cascading errors could spread before anyone notices. Oracle's current answer involves monitoring and audit tooling, but Abbabatulla remains unconvinced.

"I don't see a clear response from any vendor on the liability issue," he said, noting that this concern has been raised repeatedly in the industry over the past few years.

Despite these challenges, some industry observers are more optimistic about Oracle's positioning. Mickey North Rizza, IDC group vice-president for enterprise software, called it a "significant shift" in agentic systems that continuously complete work within enterprise software systems.

"Overall, this is a great move for Oracle positioning it as a market shaper towards the Agents as Apps. It won't be the app with the best UI that does well, but rather the agent that reliably completes outcomes that are at scale, with trust and bring sustained economic leverage," Rizza said.

The timing of Oracle's announcement aligns with growing pressure from corporate boards on technology teams to implement AI agents. Application, database, service layer, and cloud vendors are all competing for influence over enterprise AI strategy as the market for autonomous agents is expected to grow substantially.

Oracle is offering AI Agent Studio for Fusion Applications to help organizations build, connect, and run AI automation and agentic applications. This tool represents part of Oracle's strategy to provide not just the agents themselves but the infrastructure needed to create and manage them.

As enterprises consider adopting these technologies, they'll need to weigh the promised benefits of autonomous decision-making against the current limitations in data integration, the substantial implementation costs, and the unresolved questions around liability and accountability. The race to deploy AI agents is accelerating, but the technology's maturity and the frameworks for governing it are still catching up.