ServiceNow has announced AI Passport, a new authority layer meant to give enterprises runtime control over what AI agents and Model Context Protocol (MCP) servers are authorised to do at every point in a multi-agent workflow, the company said on 15 October 2026.
AI Passport, shadow AI detection and AI Cost Management are expected to be available in Q4 2026. They are not available yet. ServiceNow’s Operational Resilience solution for financial services is available today, according to the same release.
What AI Passport is
ServiceNow describes AI Passport as “a cryptographically signed credential that AI agents refer to at runtime, including environments such as NVIDIA OpenShell, to verify their authorized scope,” grounded in enterprise business context. When an agent stays in scope it can operate; when it drifts, “graduated responses can be enforced — redirecting, restricting, or reconnecting the AI agent with proper context — including kill switch capabilities at machine speed.”
The company says the design is “built on open specifications so that any governance product, identity provider, or enforcement point can issue and verify credentials for their own and third-party AI agents.” For third-party agents especially, that is meant to create “a verifiable permissions record rather than relying on the AI agent’s own ‘claims’ about its permissions.”
Each agent would receive its own passport issued by ServiceNow’s Super Intelligence Control Tower and grounded in live context through CMDB, GRC policies and workflow approvals, per the release.
NVIDIA as design partner
ServiceNow named NVIDIA as “a design and launch partner,” building enforcement on “the NVIDIA Open Agent Safety Platform, which features NVIDIA OpenShell open-source secure runtime software and the NVIDIA Sentry security framework.”
Itay Ozery of NVIDIA said: “As enterprises shift from experimenting with agents to running them in production, securing those agents demands enforcement at the infrastructure layer.” Jon Sigler, ServiceNow’s executive vice president and general manager for the AI Platform, said visibility alone is not enough and that enterprises need confidence every agent is operating within authorised scope.
Broader governance bundle
Alongside AI Passport, ServiceNow announced expanded Super Intelligence Control Tower capabilities, including:
- Shadow AI detection for unapproved agents on the network
- “400+ out-of-box connectors that will enable AI discovery”
- Cost governance and value intelligence linking spend to outcomes
- AI Cost Management in IT Asset Management (ITAM), intended to bring frontier-model, SaaS AI, hyperscaler and data-centre GPU costs into one view
For banks, ServiceNow cited its own 2026 Banking AI Maturity Index: “while 100 percent of banks report using AI, only 24 percent have the governance to match it.” That is ServiceNow’s survey figure; the release does not detail methodology. The new Operational Resilience solution for financial services is positioned to find ungoverned assets and agents and shut them down if needed, and is the one piece ServiceNow says is available today.
Customer colour in the release: LTM’s CISO Chandan Pani said LTM has “deployed 1,500 AI agents across the organization, each assigned an employee ID and manager” — a customer statement, not a ServiceNow product metric.
Why it matters
Enterprises are stacking agents and MCP tools faster than they can explain who authorised which action. A signed runtime credential with kill-switch enforcement is ServiceNow’s bid to move from after-the-fact audit logs to prevention at the edge — especially for third-party agents whose own permission claims are untrusted.
Because AI Passport is only expected in Q4 2026, today’s announcement is a product roadmap plus partner story (NVIDIA OpenShell / Sentry), not a feature customers can turn on this week. Regulators and CISOs will care whether open credential specs actually interoperate outside ServiceNow’s tower once the product ships.
What’s still unknown
- Exact Q4 2026 ship week and which ServiceNow SKUs include AI Passport
- How open the issuing/verifying specs are for non-ServiceNow identity providers on day one
- Pricing for Passport, shadow AI detection and AI Cost Management
What to watch
Watch ServiceNow’s Q4 release notes and NVIDIA GTC Berlin session SB61172 for a firmer availability date and a working OpenShell enforcement demo.
What this means
- Announced, not shipping. AI Passport is expected in Q4 2026 — not available yet.
- Runtime authority for agents and MCP. Signed credentials, graduated containment and kill switches, per ServiceNow.
- NVIDIA is a design partner. Enforcement path runs through OpenShell and Sentry in ServiceNow’s telling.
Sources
Related stories
- Raleigh’s AI Help Desk and TCS’s “US$ 3.1 Billion”: Two Enterprise AI Claims, Explained
- Workday to Acquire C1.ai, an Identity Platform for Managing What People and AI Agents Can Do
- Harbor Comes Out of Stealth With $10 Million to Measure, Control and Insure AI Agents
- Fences for AI Agents: Microsoft’s Execution Containers and AWS’s Strands Box
- The Kill Switch: Why Your AI Agents Need a Safety Net (And How to Build One)
- NVIDIA Pledges $250 Million in Computing Power for UK Research on Safer AI Agents
- New Zealand’s Central Bank and Markets Regulator Set Out What They Expect From Firms on Frontier AI and Cyber Risk
- Hong Kong’s HKMA Unveils ‘ART’ Framework for Responsible AI in Banking, With an Agentic-AI Focus
Follow @tsnmediahub on X for breaking AI news as it happens.
Follow on X
