Oracle Corp. (NYSE: ORCL) and Google Cloud announced an expansion of their strategic partnership on April 22, 2026, introducing the Oracle AI Database Agent for Gemini Enterprise. The new integration allows joint customers to interact with enterprise data stored in Oracle databases using natural language, eliminating the need for manual SQL query generation. The announcement was made during the Google Cloud Next 2026 conference, marking a significant step in the companies' efforts to operationalize generative AI across distributed cloud environments.
The Oracle AI Database Agent is now available via the Google Cloud Marketplace. It functions as a natural-language-to-SQL (NL2SQL) interface, interpreting business questions and returning answers grounded in trusted enterprise context. According to Nathan Thomas, senior vice president of product management at Oracle Cloud Infrastructure, the system applies AI directly at the database layer to improve accuracy and maintain security controls without exposing sensitive data to external models. The agent is designed to handle complex queries, such as analyzing regional revenue trends or identifying inventory risks, by leveraging Oracle’s Select AI capabilities. Satish Thomas, vice president of applied AI and platform ecosystem at Google Cloud, noted that the agent provides customers with greater flexibility to turn data stored in Oracle databases into meaningful business value through the Gemini Enterprise experience. This collaboration is particularly significant given that 97 percent of Fortune 100 companies currently utilize Oracle to manage mission-critical workloads.
A key technical feature of the partnership is the preservation of data governance. Arpan Shah, Oracle’s senior vice president of database product marketing, stated that the architecture avoids data duplication or movement into separate AI pipelines. Instead, query processing remains within the Oracle database. Security is managed through OAuth authentication, ensuring that every query runs under the individual user’s database identity. This allows for fine-grained access control at the row and column levels, meaning users only receive results from data they are authorized to view.
Beyond the new AI agent, the companies announced that Oracle AI Database@Google Cloud has expanded to 15 global regions. New locations include Tokyo, Osaka, Mumbai, Delhi, Sydney, Melbourne, Milan, Frankfurt, Montreal, Toronto, São Paulo, London, Iowa, Ashburn, and Salt Lake City. Additional regions in Turin and Mexico are scheduled for deployment within the next 12 months. The expansion is supported by new technical integrations, including OCI GoldenGate for real-time database migration and an integration between Oracle Autonomous AI Lakehouse and Google BigQuery. The latter allows Oracle systems to directly read BigQuery Iceberg tables without data replication.
Enterprise adoption of the joint service includes Worldline, a major European payment processor. Arni Smit, director of software engineering at Worldline, noted that the integration of Oracle Exadata within Google Cloud provides the low-latency and high-throughput required for global transaction processing. Oracle confirmed that the AI Database Agent will be available at no additional cost for existing Oracle Autonomous AI Database customers on Google Cloud, aiming to reduce the technical bottleneck between business users and data insights.