The rise of agentic AI has prompted a shift in cybersecurity strategies, moving protection directly into databases. Experts emphasize that traditional perimeter defenses are no longer sufficient as AI-driven attacks can bypass standard security measures. This approach ensures that access policies remain enforced regardless of the instructions given to an AI agent.
During the Oracle event titled AI Cyberattacks Are Escalating: How to Secure Your Data Now, analysts discussed an emerging shared accountability model. In this framework, vendors are responsible for the secure operation of their products, while customers must identify critical assets and define their risk tolerance. The focus is on securing data at the source.
Recovery strategies are also evolving beyond simple data restoration. Organizations now need to define a trusted operational state to restore business processes effectively after an incident. This requires a deeper understanding of how AI agents interact with systems and the ability to roll back business functions to a secure, pre-incident state.
