ServiceNow has launched AI Passport and significant advancements in AI governance, which will help ensure AI agents are ready to operate in production environments by combining runtime governance, platform integration and end-to-end monitoring, all within ServiceNow’s AI operational infrastructure.
Specifically, AI Passport introduces a new authorization layer that allows companies to control, at runtime, what actions AI agents and Model Context Protocol (MCP) servers are authorized to perform in each phase of a multi-agent workflow. For its part, ServiceNow expands its governance capabilities so that companies can identify which AI systems are in place, control the actions of agents and measure spending in relation to the results obtained.
To securely execute business workflows, AI agents need more than just a reasoning model. They need a structure that determines what data and tools they can access, clear rules about what actions they are allowed to take, and an auditable record of who has authorized those actions. ServiceNow provides the operational foundation necessary for companies to safely deploy AI agents in production environments.
The activity of AI agents is governed
AI Passport is a cryptographically signed credential that AI agents query at runtime, including in environments such as NVIDIA OpenShell, to verify the scope of their authorizations, referencing the business context to ensure their decisions conform to governance policies. As long as an AI agent acts within the limits set in its passport, it can operate freely. If he deviates from them, gradual measures can be taken, such as redirecting him, restricting his actions, or providing him with appropriate context again.
These measures include immediate deactivation mechanisms (kill switch), capable of acting at machine speed when it is necessary to stop its execution. ServiceNow’s Super Intelligence Control Tower maintains a complete audit trail of actions taken, allowing AI agent activity to be governed, auditable, and have clearly defined responsibilities.
«As AI agents multiply in companies, visibility alone is no longer enough. “Organizations need to be confident that each AI agent is operating within the boundaries of their authorizations,” said Jon Sigler, executive vice president and general manager of AI Platform at ServiceNow. “ServiceNow answers this need with an operational infrastructure that integrates AI Passport and new capabilities, unifying agent authorizations, governance and operational oversight so that companies can expand the use of agentic AI without losing control.”
AI Passport: verifiable authorizations
AI Passport is based on open specifications that allow any governance solution, identity provider, or policy enforcement mechanism to issue and verify credentials for first-party and third-party AI agents. These credentials can be verified at each stage of multi-agent workflows.
This approach represents a fundamental change, moving from reaction to prevention. Unlike audit logs, which show what happened after the fact, AI Passport helps prevent AI agents from acting outside of established boundaries by verifying at runtime what actions they are authorized to do.
In this way, it materializes ServiceNow’s Shift Zero approach, based on incorporating security by design and from the first moment. Each AI agent receives its own AI Passport, issued by Super Intelligence Control Tower, and linked to updated business context through ServiceNow’s configuration management database (CMDB), governance, risk and compliance (GRC) policies, and workflow approvals. For third-party AI agents, this allows for a verifiable record of their permissions, rather than relying on authorizations that the agents themselves claim to have.
AI Passport’s approach allows you to precisely revoke permissions and delegate authorizations with increasingly narrow scope at each stage, facilitating governance of third-party agents. It also provides audit trails with complete signatures attesting to regulatory compliance.
ServiceNow also announces NVIDIA as its design and release partner, with which it is developing a production-ready implementation of these control mechanisms through the NVIDIA Open Agent Safety Platform, which includes the open source NVIDIA OpenShell software for secure execution environments and the NVIDIA Sentry security framework.
“As companies move from experimenting with AI agents to using them in production environments, ensuring their security requires applying control mechanisms at the infrastructure level,” said Itay Ozery, director of Product Marketing for Networking and Cybersecurity at NVIDIA. “By deploying AI Passport with NVIDIA Open Agent Safety Platform, ServiceNow customers can continuously verify and enforce authorizations for each agent, at machine speed and at runtime.”
Visibility, control and measurement
In addition to AI Passport, ServiceNow offers comprehensive AI governance on a single unified platform, covering discovery, visibility, control, measurement and compliance. The new capabilities announced today build on ServiceNow’s autonomous remediation and kill switch capabilities, which enable businesses to act quickly.
ServiceNow AI Platform runs on a single, unified data model, automating governance decisions based on asset insights, identity security, and AI observability, eliminating the need to manually transfer information and tasks between teams:
Control of AI in highly regulated sectors
In financial institutions, assets without adequate governance and shadow AI multiply the risks associated with regulatory compliance and regulatory deadlines. Regulators are reducing the times available to report incidents at a rate far greater than the response capacity of manual, spreadsheet-based systems. ServiceNow’s AI in Banking Maturity Index 2026 highlights this gap: while 100% of banks say they use artificial intelligence, only 24% have governance mechanisms commensurate with that level of adoption. Unmanaged service accounts, shadow AI agents, and unregistered devices continue to represent risk exposure points.
ServiceNow provides the operational foundation necessary for companies to safely deploy AI agents in production environments
ServiceNow’s new Operational Resilience solution brings together AI governance, risk management and security in a single environment, with centralized data that helps financial institutions respond quickly. In this way, Risk Directors (CRO) and Information Security Managers (CISO) have the same visibility and control capacity in real time.
Super Intelligence Control Tower detects AI assets, identities, and agents without proper governance across information technology (IT), operational technology (OT), cloud, and third-party systems, all on a single data model and unified platform. This solution is designed to treat an AI agent that acts outside authorized boundaries like any other exposed asset: identify it, generate an alert, and, if necessary, disable it. When an incident occurs, banks can assess its severity, contain it and revoke access within the deadlines established by regulations, also having auditable evidence of the measures taken.
ServiceNow AI Passport, shadow AI detection, and AI Cost Management are expected to be available in Q4 2026. ServiceNow’s Operational Resilience solution for the financial sector is available now.
