ServiceNow adds AI agent controls and employee automation tools
SiliconANGLE Paul Gillin
ServiceNow rolled out AI agent controls and new employee automation tools. It’s trying to make workplace bots both useful and harder to let run wild.
Based on reporting by SiliconANGLE, Paul Gillin — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
ServiceNow is pushing on two fronts at once: tighter control over AI agents, and more ways for employees to spin up useful automations without touching code. The company announced AI Passport, a cryptographically signed credential meant to define what an agent is, what it can do, what systems it can reach and who owns it. The pitch is simple enough. If agents are going to move across enterprise software, they need more than a warning label.
AI Passport comes through ServiceNow’s AI Control Tower and carries permissions across agents, tools and vendors while checks happen as work is being done. ServiceNow says the credential pulls in business context from its configuration management database, governance policies and workflow approvals. In one demo, a third-party customer service agent was allowed to issue refunds under $25. An $18 refund went through automatically, while a $40 request was stopped at runtime and sent to human review.
That “dimmer switch” idea is the heart of the control story. ServiceNow says companies should be able to pause an agent by revoking session tokens, narrow its access to approved tools and endpoints, or shut it down by revoking credentials. If needed, the agent can later be restored with an explanation on record. Nvidia is part of the launch, linking AI Passport with its Open Agent Safety Platform, including OpenShell and Sentry, so policy can be enforced where the agent actually runs.
ServiceNow is also widening AI Control Tower beyond governance. It is adding shadow AI detection, more than 400 vendor connectors for discovery and monitoring, and cost tools that break spending down by user, department and runtime. AI Cost Management pulls together spending on frontier models, embedded AI services, hyperscalers and data center GPUs. A separate Operational Resilience product is aimed at financial institutions.
On the employee side, EmployeeWorks is getting a third-generation AI harness that combines reasoning, memory and links to business systems, with admin controls on what workers can build and access. ServiceNow is calling out Productivity Assistants for role-specific work, Productivity Routines for repeatable tasks, and ToolSmith for developers who want to create tools from natural-language prompts. NTT Data is already using AI Control Tower for shadow IT and is starting to adopt EmployeeWorks, but it says consolidation across tools and systems still takes time. Pricing mixes seat licenses with AI credits called assists, and the new governance tools plus AI Cost Management arrive later this quarter. The EmployeeWorks additions are in controlled availability now, along with Operational Resilience for financial services.
My take — AI-written commentary, not fact-checked reporting
This is what enterprise AI looks like once the demo smoke clears: less magic, more permissions. ServiceNow understands the real problem isn’t whether agents can do work, it’s who gets blamed when they do the wrong work at machine speed. The industry keeps selling autonomy, then immediately has to invent a leash.
Read more about this at: SiliconANGLE