Analyst(s): Brad Shimmin
Publication Date: October 20, 2025
What is Covered in this Article:
- Oracle continues to evolve its flagship database by introducing Oracle AI Database 26ai, positioning it as an “AI-native” platform.
- Oracle has launched Select AI Agent, an in-database framework for building, deploying, and managing autonomous agentic AI workflows.
- The new Oracle Autonomous AI Lakehouse makes for a direct competitor to Databricks and Snowflake, built on the open Apache Iceberg format.
- Oracle’s multi-cloud strategy for its database services aims to bring these new AI capabilities to customers on OCI, AWS, Azure, and Google Cloud.
The Event – Major Themes & Vendor Moves: At its AI World 2025 event in Las Vegas, Oracle delivered a clear and forceful message: the future of enterprise AI isn’t in a separate, bolted-on stack but is instead inextricably linked to the data—and therefore, to the database itself. The centerpiece of this strategy is the newly announced Oracle AI Database 26ai, presented not as a typical upgrade but as a seamless “transition” from the previous 23ai release via a simple release update with no re-certification required. This seemingly minor detail underscores a major strategic point. Oracle wants to make adopting its AI vision as frictionless as possible for its massive install base.
The announcements reveal a multi-pronged strategy to re-centralize the AI stack around the database. Instead of just adding vector search capabilities, which have become table stakes, Oracle is embedding the entire agentic AI lifecycle directly into the database engine with Select AI Agent. This framework allows developers to define and run multi-step agentic workflows using familiar languages like PL/SQL and Python, keeping data processing, reasoning, and action-taking within the secure, performant confines of the database.
Simultaneously, Oracle is aggressively pushing into the analytical domain with its Autonomous AI Lakehouse. By embracing the open Apache Iceberg table format, Oracle is seeking to challenge the dominance of players like Databricks directly. The value proposition delivers an open, multi-vendor data lake architecture while leveraging the performance of Exadata and the robust security of the Oracle Database. This entire vision is underpinned by Oracle’s growing importance as a key infrastructure provider for major AI players, evidenced by its deepening relationships with NVIDIA and, more recently, AMD, and its commitment to making its database services available across all major hyperscalers.
Oracle AI World 2025: Is the Database the Center of the AI Universe Again?
Analyst Take: Oracle is executing on an extension of its existing playbook, one that’s now supercharged for the generative AI era. While the market fragments into a dizzying array of specialized vector databases, MLOps platforms, and competing agentic frameworks, Oracle is making a contrarian bet on convergence. The company is arguing that the most secure, performant, and governable way to build enterprise AI is to bring the AI to the data, not the other way around. This isn’t just a product update for Oracle; it’s an architectural statement of intent designed to bring the company and its customers forward in adopting its newest database technologies.
The Converged Database is Oracle’s AI Moat
For years, Oracle has championed its converged database, which unifies relational, JSON, graph, and spatial data models. With 26ai, the company is adding vectors and AI agents as native, first-class citizens. This is Oracle’s strategic moat. Instead of stitching together a vector database for RAG, a separate relational database for transactions, and an external Python framework for agents, each with its own security model and data movement challenges, Oracle offers a single, integrated platform. For example, the ability to run a hybrid search across vectors and structured relational data in a single SQL query is a powerful simplifier that can eliminate brittle ETL hops and reduce architectural complexity.
Agents as First-Class Database Citizens, a Bold Move
The introduction of Select AI Agent is an audacious part of this announcement. While the rest of the world standardizes on external frameworks like LangChain, Oracle encourages developers to build agents *inside* the database itself. The technical rationale is sound: eliminating network latency by co-locating agent logic with the data can provide massive performance gains. Likewise, applying the database’s granular security policies directly to agent actions offers a more robust governance model. Letting millions of existing PL/SQL developers build AI agents is a clever play to activate its existing ecosystem by adding some shine to the well-traveled PL/SQL architecture. This is a direct challenge to the current architectural consensus, and its success will depend on whether the benefits of integration outweigh the allure of the broader open-source ecosystem.
The Lakehouse Gambit: Embracing Openness
Oracle’s Autonomous AI Lakehouse is a shrewd and necessary competitive response. With this release, the company acknowledges that it cannot force the entire world onto its proprietary formats. By fully embracing Apache Iceberg, it’s telling customers, “Keep your open data lake; we’ll just run it better.” The promise of combining the interoperability of Iceberg with the legendary performance of Exadata and the security of Autonomous Database is what Oracle’s George Lumpkin called a “Lakehouse platform without trade-offs.” This allows Oracle to enter conversations with Databricks and Snowflake customers not as a replacement, but as a powerful, complementary query and AI engine that can coexist in a multi-vendor environment.
The one potential catch in this grand vision remains the adoption of Oracle Autonomous Database, the foundation for these new capabilities. While technically robust, Oracle has faced headwinds in convincing its vast on-premises customer base to move to its flagship autonomous cloud service. The compelling nature of these new AI features may finally be the catalyst the company needs to accelerate that migration.
What to Watch:
- Adoption of In-Database Agents: Will enterprise development teams embrace building agents in PL/SQL and Python inside the database, or will the gravity of external open-source frameworks prove too strong? Watch for early customer wins and case studies.
- Lakehouse Performance Benchmarks: Oracle is making bold claims about Exadata’s performance on open Iceberg tables. Expect competitors and third parties to test this. Head-to-head comparisons with Databricks and Snowflake will be critical.
- Autonomous Database Migration Rate: The success of this strategy hinges on customers using Oracle Autonomous Database. Track adoption rates to see if these AI features are compelling enough to overcome migration inertia.
- Developer Tooling Maturity: Features like the planned APEX AI Application Generator and the Private Agent Factory are promising. Their usability and power will be key to winning over developers who are central to building the next generation of AI applications.
You can read the full press release at Oracle’s newsroom.
Disclosure: Futurum is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.
Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of Futurum as a whole.
Other insights from Futurum:
IBM TechXchange 2025: The Real Headliner is Data, Not AI
Neo4j GraphSummit: The OG (Original Graph) Database Player Makes a Bold Play for the Future of AI
Image Credit: Oracle
Author Information
Brad Shimmin is Vice President and Practice Lead, Data Intelligence, Analytics, & Infrastructure at Futurum. He provides strategic direction and market analysis to help organizations maximize their investments in data and analytics. Currently, Brad is focused on helping companies establish an AI-first data strategy.
With over 30 years of experience in enterprise IT and emerging technologies, Brad is a distinguished thought leader specializing in data, analytics, artificial intelligence, and enterprise software development. Consulting with Fortune 100 vendors, Brad specializes in industry thought leadership, worldwide market analysis, client development, and strategic advisory services.
Brad earned his Bachelor of Arts from Utah State University, where he graduated Magna Cum Laude. Brad lives in Longmeadow, MA, with his beautiful wife and far too many LEGO sets.
