Oracle Executives Preview Enterprise AI Strategy Ahead of Las Vegas Conference

Deploying enterprise AI at scale demands far more than access to advanced models. Organizations must establish secure pathways to their data, build dependable systems and implement oversight mechanisms for autonomous agents—all essential to converting AI pilots into tangible business value.

AI is moving fast. Threats are moving just as fast, and the race is on to build enterprise AI that's not just intelligent, but trusted, secure and, of course, ready for production.

Dave Vellante, chief analyst at theCUBE Research

Oracle Corp. is preparing for Oracle AI World 2026, scheduled for Oct. 25–28 in Las Vegas. In advance of the event, nine company executives sat down with Vellante to discuss the conference's "Data Deep Dive" programming track. Their remarks centered on Oracle's approach to trustworthy agentic AI, modernizing databases, strengthening cybersecurity and advancing application development—along with the engineering innovations meant to help organizations scale AI deployment.

The interviews preview Oracle's product demonstrations, technical sessions and interactive workshops, including an Oct. 28 hackathon pairing OpenAI Codex with Oracle AI Database. TheCUBE plans to cover Oracle AI World on-site on Oct. 28, offering further analysis of how enterprises are adopting AI technologies and the systems supporting that adoption.

Nine Insights from Oracle Leadership

1. Enterprise AI trust requires security across the entire technology stack

Organizations can only embrace agentic AI if they believe the systems will reliably handle decisions and actions tied to critical business functions, according to Juan Loaiza, executive vice president of Oracle Database Technologies. Loaiza told theCUBE that simply layering security measures onto existing systems falls short when AI applications demand consistent safeguards and dependability. His session at Oracle AI World will examine how merging AI, data management and application development into a single unified platform delivers security and governance throughout the stack. This reflects a core principle of Oracle's Data Deep Dive program: security must be built into the foundation, not delegated to individual models.

2. Oracle AI Database advances enterprise AI with new capabilities and stronger security

Connecting intelligent systems to the enterprise information that informs their answers is essential for practical AI, according to Hasan Rizvi, EVP of product and research, database engineering, at Oracle. Rizvi's upcoming presentation will detail Oracle AI Database improvements designed to streamline complex data setups, enable flexible deployment models and safeguard confidential data. He also outlined features for handling both structured and unstructured information, constructing AI agents without moving data to external systems and enabling AI-assisted application creation. Data protection remains central to Oracle's focus, with emphasis on guarding information at its source, responding swiftly to security incidents and maintaining system resilience during breaches.

3. Database-level controls establish trust for autonomous AI agents

Autonomous agents handling sensitive corporate data must follow enforceable security and business constraints, even if the applications or agents themselves become compromised, explained Tirthankar Lahiri, senior vice president of mission-critical data and artificial intelligence engines, Oracle Database, and Sara Narayan, VP of product management, mission-critical data and AI engines, Oracle Database. Their session will showcase Oracle's trust framework across four dimensions: data, access, knowledge and execution. The executives described capabilities including Trusted Answer Search, which links conversational queries to verified answers, and JSON Duality Views, which limit agent-initiated changes to databases through restricted APIs. They also introduced Oracle Database's Deep Trust Architecture, which enforces protections at the database layer regardless of whether requests originate from users, applications or autonomous agents.

4. AI-powered cyberattacks increase the urgency of database security

AI-assisted attacks are accelerating the pace at which organizations must detect flaws and roll out fixes, according to Vipin Samar, SVP of software engineering, database security, at Oracle. Samar noted that attackers can now use AI to analyze released patches and uncover the vulnerabilities they address, rendering conventional update schedules inadequate. His Oracle AI World talk will focus on expanding visibility into security risks across database environments and speeding up patch deployment before attackers can exploit weaknesses. Oracle's approach centers on protecting data where it lives, responding at automated speeds and bolstering defenses against increasingly advanced threats.

5. Autonomous database infrastructure provides a foundation for enterprise AI at scale

The distance between successful AI experiments and production systems often hinges on the quality of the underlying data infrastructure, according to Çetin Özbütün, EVP of product and research, data warehouse and autonomous database, at Oracle. Özbütün stressed that contemporary AI workloads demand automated security, guaranteed uptime and flexible resource scaling while pushing organizations to minimize operating expenses. His presentation will demonstrate how Oracle Autonomous AI Database handles enterprise AI workloads through self-managing administration, automated security patching and scalable resources. The discussion will also explore the financial benefits of database automation and why a solid data foundation is necessary to achieve real returns from AI investments.

6. Autonomous AI Lakehouse brings analytics and AI-driven decision-making together

Enterprise data lakehouses are evolving beyond their traditional role in analytics to serve as platforms for AI-informed business decisions, according to Nipun Agarwal, SVP of software engineering, Oracle Database, at Oracle. Agarwal introduced new Oracle Autonomous AI Lakehouse features intended to let organizations extract insights from data spread across multiple locations without extensive data transfers. His event talk will explain how automatically enriched business context, more precise retrieval of pertinent data and expanded Apache Iceberg compatibility can enhance AI outcomes while lowering inference expenses. The session will also highlight centralized security controls, data lineage tracking and an AI-assisted development environment designed to make analytics and AI work simpler.

7. Exadata innovations improve AI performance and database resilience

Executing AI workloads alongside critical business transactions requires infrastructure that delivers speed without sacrificing reliability, according to Jia Shi, group VP of data systems software engineering at Oracle. Shi highlighted Exadata Live Migration, which allows active database virtual machines to shift between servers using Remote Direct Memory Access while keeping connections intact and avoiding service interruptions. Her session will also discuss how Exadata accelerates AI vector searches and supports deployment on Oracle Cloud Infrastructure, competing cloud services and local data centers. A secondary theme is the application of agentic AI to pinpoint and resolve database input/output performance issues through self-managing systems.

8. APEXlang brings AI-generated applications closer to enterprise production

AI coding assistants are making software creation more accessible, yet enterprises require applications they can comprehend, update and protect, according to Mike Hichwa, SVP of software engineering at Oracle. Hichwa introduced APEXlang, a formal language for defining applications that permits AI coding assistants to build low-code systems in a structured, compilable form. The method allows tools such as OpenAI Codex and Claude to generate application definitions while the Oracle APEX platform manages execution. His talk will address how this separation of concerns can reduce accumulated technical problems, preserve human readability of AI-created code and satisfy enterprise needs for control and growth.

9. Oracle AI World puts agentic AI development into practice through hands-on sessions

Oracle AI World will give attendees the chance to move beyond theoretical discussions and work directly with tools for building production systems, according to Arpan Shah, SVP of database product marketing at Oracle. Shah described the Data Deep Dive program, which mixes executive product presentations with technical workshops, live demonstrations and sessions on database protection and agent design. A centerpiece is the Oct. 28 hackathon combining OpenAI Codex with Oracle AI Database, where participants can develop applications with support from specialists at Oracle and OpenAI. The event targets professional developers and technically engaged participants, offering hands-on experience building AI applications using enterprise database systems.

TheCUBE will deliver on-site reporting from Oracle AI World 2026 on Oct. 28 in Las Vegas. Through conversations with Oracle leaders and industry observers, the outlet will offer additional perspectives on enterprise AI developments, database technology and security advances, with attention to how organizations are putting these tools to work. Coverage will also surface key announcements and practical insights from the conference.