Oracle Empowers Fusion App Developers with Innovative Agentic AI Tools

Key Takeaways

Oracle has launched new tools for developers to build AI agents for its Fusion Applications suite.
The updates target both no-code and pro-code environments, allowing users to build applications tailored to their needs.
Challenges remain, such as ensuring security, data quality, and managing multi-agent interactions within enterprise settings.

New AI Capabilities for Oracle Fusion Applications

Oracle has introduced new capabilities aimed at enhancing the development and governance of AI agents within its Fusion Applications suite. This suite covers various sectors, including finance, supply chain, manufacturing, HR, sales, customer service, and marketing. The latest upgrade offers developer-focused tools that combine no-code and traditional pro-code options. These tools are designed to complement the pre-existing functionalities in the Oracle AI Agent Studio for Fusion Applications, first unveiled in March 2025.

The key feature of this update is the AI-native builder experience, which grants developers access to popular AI coding models and agents from third-party vendors such as OpenAI Codex, Claude Code, Microsoft VS Code, Google Antigravity, and Git. Fusion suite developers can also utilize command line interfaces from several vendors.

Oracle has branded applications developed using these tools as “Fusion Agentic Applications.” These applications operate through specialized agents capable of reasoning, coordinating, and executing tasks within Fusion workflows. Natalia Rachelson, Oracle’s Senior Vice President of Cloud Applications Development, highlighted the importance of accommodating both non-coders and professional developers with these offerings, emphasizing flexibility in environment choices for programming.

Analyst Robert Kramer views the new tools as a moderate enhancement to Oracle’s agentic platform. He notes that while they represent more than just another development tool, their effectiveness will only be gauged by actual customer outcomes. The integration of agentic AI into enterprise resource planning (ERP) systems is perceived as both valuable and risky. While simple tasks like summarizing invoices may be manageable, complex processes such as updating financial records or initiating payments come with added challenges associated with security and audit trails.

Kramer recognizes the potential for broader application of the agentic platform, extending into human capital management and customer experience. He notes the advantages of expediting production deployment and fostering collaboration among business users, developers, and partners via a range of coding options.

However, there are significant challenges to address. These include managing multi-agent failures, defining user control in application building, and ensuring data quality. Existing security measures do not resolve issues related to accountability or process ownership. Kramer calls for clearer information regarding runtime and integration costs, as well as tangible outcomes for customers.

Oracle’s shift towards AI comes amid significant strategic pivots, including the high-profile Oracle-OpenAI Stargate project aimed at establishing a robust AI data-center network. This ambitious initiative, initially projected at $500 billion, has faced challenges, particularly in Texas regarding resource management.

Oracle’s heavy investment of over $18 billion into AI data centers has raised concerns about the risks the company faces and its ability to generate sufficient revenue to meet repayment obligations. The company’s embrace of an AI-focused business model also coincides with substantial layoffs, affecting around 21,000 employees or 13% of its workforce.

As Oracle navigates these updates and challenges, its trajectory in the enterprise AI market remains a pivotal focus for stakeholders and industry observers.

The content above is a summary. For more details, see the source article.