ServiceNow is positioning its AI Control Tower as a central governance tool for enterprise AI agents. The company outlines a five-step framework for managing agent behavior: discover, observe, govern, secure, and measure.
Bhakti Pitre, vice president of AI platform security, emphasizes that managing rogue agents requires more than a simple kill switch. The company evaluates risk by mapping agents to specific business processes and the tasks they support.
The platform uses identity security technology from Veza to manage permissions. For example, if an agent authorized to provide a 10% discount is compromised to offer 100%, the system identifies the need for immediate intervention.
ServiceNow is collaborating with Okta to secure agent identities and session tokens. This partnership aims to address the complexity of agent interactions across endpoints, networks, and gateways.
