ServiceNow AI Passport Sets Hard Limits for Autonomous Agent Actions

ServiceNow Inc. announced new governance controls for artificial intelligence agents alongside updates to its EmployeeWorks portal. These changes allow employees to create specialized assistants and automate routine tasks without coding, while ensuring those tools operate within approved enterprise boundaries. The announcement addresses the challenge of expanding autonomous AI usage while maintaining strict operational control across business systems.

By the numbers

• $25: The maximum refund amount a demonstrated third-party agent could process automatically.
• $18: The specific refund amount that proceeded automatically during the vendor demonstration.
• $40: The refund request rejected at runtime because it exceeded the human review threshold.

What the figures mean

The demonstration figures illustrate how AI Passport enforces authority limits in real time. An $18 refund was allowed because it fell under the $25 threshold, while a $40 request was blocked for requiring human review. This shows the system actively checks permissions as work executes, rather than relying on advisory guardrails. ServiceNow states that identifying agents is insufficient without these enforceable limits on their authority.

What should you do?

Managed service providers should verify if their clients use ServiceNow’s AI features and assess current agent permissions. Check whether existing automation tools have defined boundaries or operate with broad system access. Review vendor contracts to understand liability for actions taken by autonomous agents. Ensure that incident response plans include steps for pausing or revoking agent credentials during security events.

Operational impact

Teams usually need to map which AI agents interact with critical infrastructure and who owns each credential. Check whether current governance policies cover third-party tools and how revocations are communicated across departments. Inform security and compliance leaders about any agents with write access to core systems. Ensure that audit logs capture both automated actions and manual overrides for accountability during investigations.

Background: Incident response plans

Incident response plans fail when teams assume they cover every attack vector or treat them as static documents. You must focus on decision trees, not scripts. Plan for communication breakdowns, not just technical fixes. Regular table-top exercises reveal gaps that documentation hides.

Read the full guide: Fix Incident Response Plans That Fail Under Pressure

What to do and how to stay safe: ServiceNow

• Audit all active AI agents to confirm they have defined identity credentials and permission scopes.
• Verify that runtime checks are enabled to block actions exceeding predefined thresholds or policies.
• Establish a process for revoking session tokens or deactivating agents that exceed their authority.
• Monitor integration points with infrastructure platforms to ensure enterprise policies are enforced consistently.

Step-by-step guide: Shadow AI Explained: 8 Questions Answered for Secure Operations

General security guidance from the BitShield News newsroom. It is not confirmed advice from the organisations named in this story.

Frequently asked questions

What is ServiceNow AI Passport?

It is a cryptographically signed credential that defines an AI agent’s identity, permitted actions, accessible systems and accountable owner.

How does AI Passport enforce limits?

It draws on business context from configuration management databases and governance policies to establish enforceable limits on agent authority.

Which company partnered with ServiceNow on this launch?

Nvidia Corp. is a design and launch partner, linking AI Passport with its Open Agent Safety Platform and Sentry security framework.