Oracle and Google Cloud expanded their partnership to bring Google Gemini models into Oracle enterprise applications, giving customers another model option for AI agents and embedded automation.
The companies said Gemini models are planned for Oracle AI Agent Studio for Fusion Applications, a development platform for building and running AI automation and agentic applications. Oracle also plans to use Gemini models for embedded AI use cases in Oracle Fusion Applications and NetSuite.
More Model Choice Inside Business Workflows
For enterprise application customers, the announcement is less about access to a single model and more about where AI execution happens. Oracle is positioning Gemini as another model option inside systems used for finance, HR, supply chain, customer experience and ERP workflows.
That matters because many enterprise AI projects stall when model output has to be moved manually into business systems. Oracle said Fusion Applications can turn agent reasoning into governed workflows, approvals and transactions, while NetSuite will evaluate Gemini for use cases where the model can improve visibility, automate work and move users from insight to action.
Agentic Applications Need Governance
The partnership builds on customers’ existing access to Gemini models through Oracle Cloud Infrastructure Enterprise AI and Google Cloud’s Gemini Enterprise Agent Platform integration. Oracle said customers and partners will be able to use Gemini 3.1 Flash-Lite for price-performance-focused use cases and Gemini 3.5 Flash for more complex reasoning and specialized tasks.
The broader question for CIOs is how model choice, application governance and transaction controls fit together. Giving users more AI options inside business applications can improve automation, but it also increases the importance of approval rules, data boundaries, auditability and model selection by use case.
The Bottom Line
Oracle’s Gemini expansion gives enterprise application customers another route to apply generative AI inside operational systems rather than separate pilots. The practical value will depend on whether organizations can connect model reasoning to governed workflows without creating unmanaged AI activity across core business processes.