Oracle has released a new version of its long-term support database, called AI Database 26ai, which focuses heavily on making AI more agent-centric for businesses. This update introduces features that help developers create, deploy, and manage AI agents directly inside the database, paving the way for smarter automation in enterprise workflows. It replaces the earlier Database 23c, which was renamed to Database 23ai after adding AI-driven app development features like vector search.

The key addition is the Select AI Agent, a framework that builds on Oracle’s existing natural language data querying tool, Select AI. While the name might suggest a simple feature, it actually provides a full platform for building and managing AI agents. Developers can use pre-made tools inside the database or connect to external systems through REST APIs and MCP servers. Oracle emphasizes that this makes AI agents a first-class part of the database ecosystem.

Empowering Developers and Non-Technical Users

Oracle isn’t just targeting tech-savvy developers with these new tools. It also introduced the AI Private Agent Factory, designed for platform teams and security professionals who want to deploy private, containerized AI agents without needing coding skills. This no-code approach offers a secure way to set up AI workflows within controlled environments. Experts compare this to similar offerings from Snowflake or Google’s ADK, which help manage AI in the cloud or on-premises.

The new features are aimed at enabling enterprises to automate tasks more efficiently. With support from MCP servers, users can build agents that interact with data inside the database to answer questions or trigger actions. This tight integration means workflows can be more dynamic and responsive, especially as AI agents handle routine or complex tasks.

Other Features and Security Enhancements

Beyond AI agents, the database continues to support advanced data capabilities like vector search, data annotations, and integration with Nvidia’s NeMo microservices. It also supports Exadata, Oracle’s high-performance database platform. Security is a priority too, with features like SQL Firewall to block malicious queries, Zero Data Loss Cloud Protect to prevent data loss from ransomware, and quantum algorithms for encrypting data both in transit and at rest.

Existing users of Database 23ai can upgrade easily by applying a software update in October 2025, without needing to do a full database overhaul. This makes the transition smooth and straightforward. The new version also serves as the foundation for Oracle’s Autonomous AI Lakehouse, which uses Apache Iceberg to ensure compatibility and flexible data management.

In summary, Oracle’s latest database release is a big step toward integrating AI agents deeply into enterprise data environments. It offers tools for developers and platform teams alike, making automation smarter and more secure. As AI continues to evolve, Oracle’s focus on agent-driven workflows could change how businesses handle data and automation tasks.

Inspired by

• https://www.infoworld.com/article/4072128/oracle-targets-agentic-use-cases-with-ai-database-26ai.html

Sources

• gmpg.org
• cio.com
• cio.com
• csoonline.com
• podigee.com