What's New for Oracle AI Data Platform

Here’s an overview of new features and enhancements in Oracle AI Data Platform.

Topics

For the latest Oracle AI Data Platform news, product updates, customer and partner stories, and market insights, see https://blogs.oracle.com/ai-data-platform/.

September 25, 2026

Feature Description
Deploy Models Models in AI Data Platform can now be deployed for use as endpoints. Deployed models can be created and managed from the new Deployments tab on your models. See Models.
Preview: Track Knowledge Base Lineage AI Data Platform Lineage now lets you visually trace the tables and volumes that feed data into your knowledge bases. See Lineage (Preview).
Ingest and use synonyms in your catalogs You can now take synonyms from Oracle Autonomous AI Database-backed data sources and use them in AI Data Platform. Synonyms can be called in notebooks and other SQL queries in place of their underlying data. See Synonyms.
External catalog support for Snowflake AI Data Platform now supports Snowflake as an external source for catalogs in your Master Catalog. See External Catalogs.
Tables now support Apache Iceberg table format You can now use Apache Iceberg as a transactional table format for tables in AI Data Platform. See Tables.
Use widgets to parameterize notebooks Widgets let you define input values in your AI Data Platform notebooks, enabling users to run the same notebook with different values, letting you reuse the notebook without editing its code each time. See Widgets in Notebooks.

August 10, 2026

Feature Description
Preview: Track the relations of data artifacts through Lineage Lineage lets you visually trace the relationships between data artifacts in your AI Data Platform instance. You can choose a specific artifact and view upstream and downstream relations. See Lineage (Preview).

July 6, 2026

Feature Description
Create AI Agents to help develop and manage your data You can now create AI agents in AI Data Platform, either through the visual flow builder or through code. See AI Agents.
Power your Agents with AI Compute AI agents require specialized compute to power their tools and skills. You can create AI compute through Compute in your AI Data Platform instance. See AI Compute.
Store documents and search documents in Knowledge Bases Knowledge bases let your store and search documents in AI Data Platform, using Oracle Database 26ai Vector Search capability to store vector embeddings. AI agents can use these knowledge bases to perform semantic searches and retrieve relevant data and documents. See Knowledge Bases.

June 5, 2026

Feature Description
Connect to data in your AI Data Platform from your IDE The Oracle AI Data Platform (AIDP) Developer Extension is now available to install from the Microsoft Visual Studio Code marketplace. Browse workspaces and catalogs, remotely preview, compare, download and upload files, and more. See Oracle AI Data Platform (AIDP) Developer Extension.

May 1, 2026

Feature Description
Choose customer-managed key encryption When creating your AI Data Platform instance you can now choose to use Oracle-managed key encryption or use your organization's managed key encryption. See Storage Encryption Options.
Enable network source-based access control Create sets of defined IP addresses that you can reference in policies or authentication settings to control access. See Network Sources.
AI Data Platform REST API Reference Documentation for AI Data Platform REST APIs is now available. See REST API for Oracle AI Data Platform.
Preview: Manage credentials in with secrets and vault references The Credential Store provides a secure and centralized way to manage sensitive information such as authentication tokens and external credentials. Instead of embedding secrets directly in configurations or code, users can store and reuse them safely through their AI Data Platform. See About Credential Store.
Preview: Connect and manage your data in Git You can now manage your Git files entirely from within AI Data Platform, with utilities that allow you to push, pull, commit, and merge files as well as manage any merge conflicts that arise without having to leave the platform. See Git Integration (Preview).
Preview: Package and deploy bundles from Git folders Bundles allow you to package resources in a Git folder, like those requires for jobs or agents, and deploy those packages across workspaces and environments to share those resources. See Bundles and CI/CD (Preview)
Preview: Run and iterate experiments to build models based on your data Experiments in AI Data Platform provide means for data scientists, machine-learning engineers, and managers to collaborate in the development of models. See Experiments (Preview).
Preview: Create and train machine-learning models Machine-learning models in AI Data Platform enable you to build optimized models through performance analysis, collaboration, and analysis of experimental conditions like hyper-parameters, input datasets, and feature engineering. See Models (Preview).
Preview: Administration Settings Administration settings let manage your linked Git accounts. See Administration Settings (Preview).