Enterprise data has long sat behind SQL barriers. Oracle just tore them down inside Google Cloud. On April 22, the companies unveiled the Oracle AI Database Agent for Gemini Enterprise, letting users query Oracle databases in plain English. No code. No schemas. Just answers.
Customers running Oracle AI Database@Google Cloud can now tap Gemini Enterprise to ask about revenue trends or operational metrics. The agent interprets the request, pulls governed data from the database layer, and spits out context-aware insights—all without shifting a byte outside secure boundaries. Nathan Thomas, Oracle’s senior vice president of product management for Cloud Infrastructure, put it this way: “We’re making it easier for customers to use natural language to access, understand, and act on enterprise data by combining the Gemini Enterprise experience with Oracle’s industry-leading database performance, security, and governance.” (Oracle announcement)
Satish Thomas—note the shared surname, no relation—of Google Cloud added: “To deliver real impact from agentic AI, customers need a simple, trusted way to interact with their valuable business data using intelligent agents such as the Oracle AI Database Agent.” (Oracle announcement) This isn’t some bolt-on chatbot. Processing happens in-database. Security holds firm with row- and column-level controls. User identities propagate, so you see only what you’re cleared for.
Worldline, the European payments giant handling billions of transactions, already leans on Oracle Exadata Database Service within Google Cloud for low-latency processing at scale. Arni Smit, Worldline’s director of software engineering for integration and payments, says it delivers “the scalability, resilience, and security capabilities we require to support real-time transaction processing at global scale.” (Oracle announcement) Japan’s AI Shift, a CyberAgent unit building AI for marketing and service, plans to roll it out too. CEO Yuto Yoneyama calls it “an important advancement,” speeding decisions without custom tools or data copies. (Oracle announcement)
This builds on groundwork laid in September 2024, when Oracle Database@Google Cloud hit general availability in regions like U.S. East (Ashburn) and Germany Central (Frankfurt). That move brought Exadata, Autonomous Database, and Zero Data Loss Recovery straight into Google’s data centers—co-located for minimal latency, no more interconnect hops. Oracle’s Karan Batta boasted back then: “I think the world’s data actually runs on Oracle databases.” Google Cloud’s Amit Zavery agreed customers crave choice: “A lot of enterprise applications have Oracle Database at the back end.” (VentureBeat, Sep. 9, 2024)
Fast-forward. Availability now spans 15 regions, from Tokyo to São Paulo, with Turin and Mexico City coming in 12 months. New tricks include OCI GoldenGate for real-time migrations—GA soon—and Autonomous AI Lakehouse tying into BigQuery for analytics without duplication. The agent lives in Google Cloud Marketplace, free for existing Autonomous AI Database users on Google Cloud, though rollout starts limited. (SiliconANGLE, Apr. 22, 2026)
Arpan Shah, Oracle’s senior VP of database product marketing, nailed the pain point: Business users ping IT for every report. “With the announcement, business users will be able to ask questions right through Gemini Enterprise in natural language.” And crucially: “There’s no data moved; it’s all done securely.” (SiliconANGLE) Developers get more: Hook it into Vertex AI or Gemini’s agent platform for extraction, visualization, even agent-to-agent chains.
Why now? Enterprises hoard Oracle data—think ERP, finance, supply chains. AI hungers for it, but moving data risks compliance nightmares. This keeps everything in place while feeding Gemini. Oracle’s multicloud push mirrors deals with AWS and Azure; multicloud database revenue rocketed 1,529% in one recent quarter. (VentureBeat, Sep. 9, 2025) Google gains Oracle’s installed base, juicing its AI stack without rebuilding.
But.
Challenges lurk. Latency spikes in cross-region setups. Token costs could climb for heavy natural-language parsing, though in-database shifts some load. Early access means waiting lists. Still, it plugs right into Gemini Enterprise Agent Platform, where partners like Atlassian and ServiceNow build agents too. (The Register, Apr. 22, 2026)
Oracle stock? Up on AI tailwinds, with OCI revenue surging. This cements its pivot: From legacy database king to AI enabler across clouds. Google Cloud, still chasing AWS and Azure, scores enterprise stickiness. Customers win biggest—query your data empire. In English.
WebProNews is an iEntry Publication