Oracle executives and industry experts discussed the security challenges of agentic AI during the virtual event AI Cyberattacks Are Escalating: How to Secure Your Data Now. The discussion highlighted the need for architectural shifts in how enterprises manage access and data protection as software agents gain the ability to act on behalf of users.
Oracle is moving security enforcement from individual applications directly into the database layer. Juan Loaiza, executive vice president of Oracle Database Technologies, noted that this approach ensures end users and their agents can only access authorized data regardless of the application used. The company integrates these controls with tools like Oracle Data Safe and Oracle Database Security Central.
Accenture Enkitec Group implemented Oracle Deep Data Security to manage access rules for agent-based applications. Roger Cornejo from Accenture reported that the system has been tested to scale to 300 simultaneous queries, with most results returned in 15 seconds or less. This setup allows agents to query shared knowledge repositories while maintaining strict authorization levels.
Experts emphasized the importance of identity management and system modernization. Johnnie Konstantas of Oracle highlighted that agents act as proxies for individuals, requiring uniform identity controls. Meanwhile, Ashish Ray noted that security must cover both production systems and backups, with the Zero Data Loss Recovery Appliance used to validate backup integrity and recoverability.
