Rubrik has released new findings from its Zero Labs research, revealing that enterprises are rapidly deploying AI agents without the necessary safeguards to manage and secure them effectively. The highlights a widening gap between innovation and security as organizations accelerate the use of autonomous systems.
Based on a of more than 1,600 IT and security leaders, the shows that 86% of respondents expect AI agents to outpace their organization’s security controls within the next year. At the same time, only 23% having full visibility into the AI agents operating in their environments, a figure the suggests may be overly optimistic. This lack of visibility creates significant challenges in managing identities that are actively making decisions, accessing data, and executing actions.
A key concern identified in the is the rapid growth of non-human identities linked to AI agents. These identities are expanding faster than organizations can track or govern them, forming what researchers describe as a “shadow workforce.” Often operating with persistent access and minimal oversight, these identities introduce new risks, including unauthorized activity, lateral movement, and potential compromise.
The also raises questions about the operational efficiency of AI agents. While organizations are investing heavily in automation, more than 80% of respondents said AI agents require more manual oversight than the efficiency gains they deliver. Additionally, 88% indicated they lack the ability to roll back agent-driven actions without causing disruption, highlighting a critical gap in control and recovery capabilities.
Concerns around resilience are also growing. Nearly nine in ten leaders expressed uncertainty about their ability to meet recovery objectives as AI-driven threats increase. The research suggests that autonomous systems are not only transforming enterprise operations but also reshaping the threat landscape, with nearly half of respondents expecting agentic systems to drive the majority of cyberattacks in the coming year.
Kavitha Mariappan, Chief Transformation Officer at Rubrik, said organizations are struggling to maintain control over systems they cannot fully observe or govern. She emphasized that as decision-making shifts from humans to machines, maintaining operational safety becomes a central challenge for enterprise leaders. Industry experts also point to identity management as a critical factor in securing AI systems. Steven Ramirez, Chief Information Security and Technology Officer at Renown Health, noted that visibility and verification of identities are essential to achieving secure and scalable AI adoption without introducing new risks.
Recommended Cyber Technology News:
- IMF Chief Flags AI Cyber Risks to Global Monetary System
- ATIS Launches 5G Zero Trust Policy Framework for Enhanced Security
- Datacom Reveals Cyber Recovery Gaps in New Zealand Firms
To participate in our interviews, please write to our CyberTech Media Room at info@intentamplify.com
