Oracle Launches Managed MCP Service for Governed AI Agent Database Access

Oracle DevelopersOracle Developers

· Updated

Oracle launched the OCI Managed MCP Service, a cloud-native gateway that connects AI agents to Oracle databases using the Model Context Protocol. This move shifts agentic data access from local developer tools to a managed enterprise environment with centralized identity and governance controls.

Oracle, the enterprise database and cloud infrastructure provider, launched the OCI Managed MCP Service to connect AI agents to cloud-hosted databases. It uses the Model Context Protocol (MCP) — a universal connector standard — to provide a managed HTTPS gateway. This replaces the local setup previously required for agentic database interaction.
Availability
OCI Database Tools Service
Protocol
Model Context Protocol (MCP) over HTTPS
Supported Databases
Oracle AI Database 26ai, Oracle Database 19c
Cloud Support
OCI, AWS, Azure, Google Cloud
Authentication
OAuth 2.0, OCI Identity (IAM)

This managed approach follows the release of Tableau's analytics platform MCP server, marking a shift toward governed agentic access for enterprise data. By moving to a managed service, Oracle addresses the security gap that local MCP tools create. It uses OCI Identity to ensure only authorized agents can query sensitive production data.

You can now define SQL Reports — pre-validated queries that agents execute via natural language without generating their own code. This reduces hallucinations and ensures accuracy. The service is available at no extra cost for OCI users and supports Oracle databases on AWS, Azure, and Google Cloud.

View the full update on blogs.oracle.com
Oracle Developers
Oracle Developers
@OracleDevs
X

Let your business users across your organization adopt AI! OCI Managed MCP Service helps by enabling AI agents and assistants easily and securely connect to Oracle databases running in the cloud using natural language to access validated SQL Reports. @krisrice and @thatjeffsmith share the details: https://t.co/BQ3ZneN8HW

3retweets7likes
View on X

Still wondering? A few quick answers below.

This is a cloud-native managed service that uses the Model Context Protocol to connect AI agents to Oracle databases. Unlike previous local versions, this service provides a secure HTTPS gateway within the Oracle Cloud Infrastructure. It allows agents to discover data, run SQL, and access validated reports using natural language while maintaining enterprise-grade security and identity controls.

Security is managed through OCI Identity and Access Management using OAuth 2.0 and federated user accounts. Administrators can assign specific roles, such as administrator, operator, or user, to control which individuals or agents can access particular databases. This ensures that agentic access follows established enterprise authorization policies rather than relying on shared or local database credentials.

SQL Reports are pre-validated, parameterized queries defined by administrators that AI agents can execute as tools. By using these reports, organizations can provide business users with accurate data through natural language without requiring the AI to generate its own SQL code. This mechanism significantly reduces the risk of hallucinations and ensures that agents return approved, consistent results.

The service supports Oracle AI Database 26ai and Oracle Database 19c instances. It is compatible with various deployment models, including Autonomous Database and Exadata Database Service. Beyond OCI, the service also provides agentic access to Oracle databases running on other major cloud platforms, specifically Oracle AI Database at AWS, Azure, and Google Cloud.

Oracle provides the native MCP Server capability within the OCI Database Tools Service at no incremental charge. Customers do not pay a separate fee for the MCP access layer itself. Instead, they only pay for the underlying database services, the large language models they choose to use, and the standard consumption costs associated with running database queries and operations.

Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →

Share this update