Research shows that 90 percent of IT decision-makers find their identity management insufficient for increasing AI risks.
Research by IDC, sponsored by Commvault, shows that 90 percent of respondents indicate that identity management must be improved to mitigate the risks of agentic AI. The results demonstrate that existing identity management systems, primarily focused on human users, are insufficiently prepared to manage a growing number of AI agents and machine identities. This increases the risks to corporate data and processes.
By investing in ResOps, companies can increase their resilience against the growing risks of AI agents and cyber incidents.
Shortcomings in identity management
The study shows that only 26.7 percent of those surveyed use dynamic, role-based access control for AI identities and data analysis. Only a quarter have established and tested procedures for protecting, detecting, and recovering Active Directory and Entra ID environments. This indicates that many organizations still need to take steps to adapt their identity management to the AI-driven reality.
Collaboration between IT and security teams also leaves much to be desired. Almost all respondents (98.4 percent) believe that this collaboration could be better, with about half indicating that substantial improvements are needed.
Resilience Operations
The IDC report points to a broader challenge: more than half of organizations have not yet defined their minimum viable business. Recovery orchestration and cleanroom capabilities are also often insufficiently developed. IDC therefore sees a growing need for Resilience Operations (ResOps), a discipline that brings different teams together to ensure business continuity and recover quickly after incidents.