Oracle adds Google Gemini to the agent menu

Oracle has announced plans to integrate Google's Gemini models into its AI Agent Studio for Fusion Applications, broadening its collaboration with Google Cloud and providing customers with additional options for building AI agents. The move aims to make it simpler for organizations to utilize Gemini models in applications and agentic workflows designed to automate processes, expedite decision-making, and achieve desired outcomes.

Google Cloud VP Satish Thomas stated that the partnership is intended to facilitate customers and partners in constructing Fusion-native agents and agentic applications within Oracle AI Agent Studio. The integration also enhances multimodal capabilities, with Gemini 3.1 Flash Lite and Gemini 3.5 Flash models available alongside Oracle's existing model choices.

Oracle's VP of applications development, Chris Leone, remarked that the addition would grant customers and partners increased flexibility when creating and enhancing agents and agentic applications that tackle intricate, real-world business complexities. Initially launching its platform for deploying LLM-powered agents within its Fusion application landscape, Oracle provided customers the opportunity to utilize models from Cohere and Meta, with the option to connect additional supported models through the platform.

However, Gartner's Balaji Abbabatulla cautioned that while the technology sounds promising, there are underlying concerns that need to be addressed. He highlighted potential challenges in implementation and questioned the responsibility for AI decision-making should an agent make an error at scale. Oracle's response thus far has been to rely on monitoring and audit tooling, but Abbabatulla remains unconvinced regarding the clarity of liability answers from vendors.