77% of IT managers say their AI agents are out of control - 5 ways to rein in yours

"Only 23% of IT managers report having complete control over the agents within their organizations, indicating a significant gap in management and oversight."

"A majority, 81%, of IT managers find that the agents require more time in manual auditing and monitoring than they were intended to save, raising concerns about productivity."

"The proliferation of unsanctioned AI applications is a growing issue, as users often bypass security controls to create agents, leading to potential security vulnerabilities."

"Agent sprawl is becoming a pervasive problem, resembling early cloud adoption, where teams independently spin up agents using different frameworks and vendors, resulting in fragmentation."

A survey by Rubrik ZeroLabs reveals that only 23% of IT managers have complete control over AI agents, with 81% reporting increased manual auditing needs. Security measures are deemed inadequate, and many users bypass controls to create unsanctioned agents. This agent sprawl mirrors early cloud adoption, resulting in fragmentation and security gaps. IT managers express a disconnect between their perceived control and the operational reality, with 86% expecting agentic productivity to decline in the near future.

Read at ZDNET