ServiceNow has introduced AI Passport, an authority layer that gives enterprises runtime control over what AI agents and MCP servers are allowed to do in multi-agent workflows. Each agent receives a cryptographically signed credential, issued by ServiceNow's Super Intelligence Control Tower and grounded in the enterprise's live business context, including CMDB records, GRC policies, and workflow approvals. At runtime, agents refer to the passport to verify their authorized scope; if they stay within it, they operate freely, and if they drift, graduated responses can redirect, restrict, or reconnect them, with kill-switch capabilities at machine speed.
Unlike audit trails, which show what happened after the fact, AI Passport is designed to prevent out-of-scope actions by authorizing them at runtime. For third-party agents, this creates a verifiable permissions record rather than relying on the agent's own claims. ServiceNow is building the solution on open specifications so other governance products, identity providers, and enforcement points can issue and verify credentials. NVIDIA is a design and launch partner, implementing AI Passport with the NVIDIA Open Agent Safety Platform, which includes the OpenShell secure runtime and Sentry security framework.
ServiceNow also announced broader governance capabilities on its unified platform, including shadow AI detection, 400+ out-of-box connectors for discovery, access mapping, and continuous observability. New cost governance features track user, department, and runtime spending, connect consumption to realized value, and integrate AI cost management into IT Asset Management. Policy enforcement manages alert-to-throttle-to-block graduated responses. The company frames these as an operational harness that combines runtime governance, platform integration, and oversight to make AI agents production-ready.