Oracle adds AI data platform tools for enterprise governance

Oracle has introduced new Oracle AI Data Platform features for enterprise AI, aimed at helping customers build, deploy and scale AI systems using shared business context.

The update focuses on tools designed to give AI agents a common understanding of business entities, relationships, policies and workflows. The platform is intended to help organisations apply governance controls, monitor usage and costs, and manage security settings as they expand AI across the business.

T.K. Anand, Executive Vice President at Oracle, said there is growing demand for AI systems that reflect how businesses actually operate, rather than relying only on models.

"AI agents are moving beyond information retrieval to participate in business decisions and execution," said T.K. Anand, Executive Vice President at Oracle. "To scale them reliably, enterprises need a system of context that reflects how the business actually works. Oracle AI Data Platform starts with business knowledge customers already have in their applications, databases, analytics, and documents and makes that context reusable across agents."

Ontology tools

A central part of the announcement is a set of business ontology tools. These ontologies define concepts, terminology, relationships and policies so analytics systems, AI agents and enterprise applications can interpret data consistently.

The platform uses what Oracle calls an AI Discovery agent to generate those ontologies by analysing database schemas, semantic models, application and API documentation, and business knowledge stored in unstructured documents. Business domain experts are then expected to review and refine the ontologies before approving them for wider use.

The system can also monitor how agents use the ontologies and recommend updates as organisational needs change. Oracle said this is intended to reduce the manual modelling work usually associated with ontology projects while keeping review and approval in the hands of business and technical teams.

Oracle also outlined the role of its Semantic Reasoning Engine, which uses the ontologies as a source of business knowledge. The engine interprets and assembles context before information reaches an AI model, handling retrieval across federated systems, resolving business relationships and applying governance controls.

According to Oracle, this setup is intended to reduce repetitive reasoning across queries and lower token use by reusing business context and prior analytical findings. That focus reflects growing scrutiny among enterprise buyers over the cost of scaling large language model-based systems.

Agent management

The announcement also included new management and governance functions for AI agents. Organisations rolling out large numbers of agents often face duplicated effort, unclear ownership and inconsistent controls, and Oracle positioned the platform as a way to address those issues through a common framework.

At the centre of that framework is Oracle AI Registry, which maintains a central inventory of agents, tools, skills and MCP servers, including ownership, versions, dependencies, permissions and lifecycle status. Oracle said business users will be able to discover and use registered agents through Oracle Agent Hub or supported third-party AI assistants, including ChatGPT and Claude.

Some governance functions remain in development. Oracle AI Governance and Control is planned to provide testing, policy assessment and monitoring of agent behaviour, quality, performance, usage and cost.

Industry view

Industry analysts cited by Oracle framed the move as part of a broader shift in enterprise AI, from model access to business-specific context and governance.

"The next competitive advantage in enterprise AI won't come from models alone. It will come from how well AI understands the business," said Megha Kumar, Research Vice President at IDC. "Oracle AI Data Platform marks an important evolution by helping organizations connect their data, business knowledge, and AI capabilities to move beyond experimentation and toward intelligent, governed business outcomes."

Kevin Petrie, Vice President of Research at BARC, also pointed to the role of proprietary data and processes in differentiating enterprise deployments.

"Anyone can adopt frontier models and agentic AI. Organisations only gain a sustainable competitive advantage when they apply those powerful algorithms to their proprietary processes and datasets," said Kevin Petrie, Vice President of Research at BARC. "The Oracle AI Data Platform offers a strategic step in that direction by helping enterprises derive governed, consistent, and reusable context from their data. Ontologies in particular enable models and agents to understand the interrelationships of key concepts and entities without consuming unnecessary tokens."

Service launch

Alongside the product changes, Oracle introduced Oracle AI Data Platform Innovation Service. The service is intended to help customers move AI projects from concept to production, with support from Oracle teams in defining business outcomes, designing systems, validating deployments and operating them in production.

Oracle said the service will use Oracle AI Data Platform and Oracle Cloud Infrastructure as the basis for those projects. It presented the offering as a way to combine technology delivery with engineering and operational support for customers seeking to put AI systems into regular business use.

The new Oracle AI Data Platform features are expected to become available within the next 12 months.