AI agents create identity governance challenges, impacting cyber resilience

Organizations are facing significant identity management challenges as they rapidly deploy agentic AI systems, according to a new survey. The systems designed to manage human identities are struggling to cope with the influx of machine identities, and many businesses have not yet defined their recovery strategies post-attack, based on information published by Channel Insider.

A survey sponsored by Commvault and conducted by IDC reveals that 90% of IT and resilience decision-makers believe their identity management capabilities need improvement to address risks from agentic AI. Non-human AI agents, with their always-on access and ability to multiply, are outpacing human users, overwhelming traditional identity systems. Over half of surveyed North American organizations (58.7%) require substantial changes to their identity management approach, with only 26.7% having implemented dynamic role-based access controls for AI. Furthermore, 57.7% of organizations have not fully defined their minimum viable business operations for post-attack recovery.

Gaps also exist in recovery orchestration, cleanroom deployments, and broader cyber resilience capabilities, with only 26% utilizing a unified backup platform and 35.4% automating the identification of clean recovery points. IDC recommends adopting resilience operations (ResOps) as a mainstream discipline to integrate business continuity, cybersecurity, and data protection for a unified recovery strategy, predicting its broader adoption in the next three to five years.

Source: Channel Insider