Oracle to Press Data-Layer Security Strategy at Sept. 22 Virtual Event on AI Cyberattacks

According to SiliconANGLE, the event is organized around Oracle's position that AI agents acting at machine speed have opened new routes into sensitive enterprise data and that policy enforcement has to sit where the data resides. Oracle first outlined the approach in June as a three-pronged strategy: secure at source, secure at speed and secure through resilience. TheCUBE will conduct interviews during the virtual event.

"The cloud shared responsibility model is no longer a sufficient framework in this AI-first world," said Dave Vellante of theCUBE Research. "Agents acting at machine speeds require what theCUBE Research calls a shared accountability model. It's not enough to protect infrastructure, apps and the data inside. AI has raised the trust bar, and customers must now consider much more deeply how adversaries using AI, and even accidental AI actions running at machine speeds bring new risks to enterprises."

The June strategy was accompanied by tools aimed at database security, patching, testing and lifecycle management. They include Oracle Deep Data Security, which applies end-user-specific privacy rules inside the database, and Oracle SQL Firewall, intended to block SQL injection attacks that let attackers run malicious database commands through website input fields. In April, Oracle announced a set of enhancements to its Oracle AI Database, among them Deep Data Security for centralized, declarative and fine-grained authorization and data visibility policies based on each end user's identity, roles and context.

By embedding Deep Data Security's policy enforcement directly in the database, Oracle says it adds protection against unauthorized data access that results from adversarially injected queries, a control the company argues becomes more necessary as agents move from answering questions to taking action. "The next phase of enterprise AI adoption will depend as much on governance as model capability," said Krista Case of theCUBE Research. "As agents gain access to sensitive data and critical workflows, organizations need to know whose identity they are acting under, what privileges they inherit and where those privileges are enforced. Getting those controls right can accelerate AI adoption."

Resilience accounts for the third part of the strategy. Oracle's Zero Data Loss Recovery products are designed to protect Oracle databases down to the last committed transaction, while the Globally Distributed AI Database uses Raft-based replication and automated failover so applications can keep running when infrastructure components, availability zones or entire sites go offline. The company has framed the AI database as the center of agentic workloads; theCUBE Research's John Furrier has described the underlying bet as one that the future of AI will be decided less by agents themselves than by where and how they interact with data.

"The resilience challenge is not only recovering from an attack," Case said. "Organizations also need to recover from authorized agents that make the wrong decisions, alter critical data or disrupt business processes at machine speed."

The Sept. 22 session is scheduled to cover the evolving threat landscape along with guidance on protecting sensitive data, limiting disruption, recovering from an attack and keeping protections current as AI-powered attacks change. Oracle executives and outside industry experts are expected to join theCUBE Research analysts for the discussion.