Submit a ticket to the IT service desk and the average response takes hours. But what if AI handled it 99% faster? ServiceNow's Autonomous Workforce, unveiled with the March 2026 Zurich release, is a paradigm shift from treating AI as a "tool" to deploying it as a "team member." It's not just slapping a chatbot on top — it's about integrating AI with defined roles like L1 Service Desk AI Specialist and Security Operations Analyst into your team.
What is this?
The hottest keyword in the enterprise AI market right now is "agents." Salesforce has Agentforce, Microsoft has Copilot Studio, and ServiceNow launched Autonomous Workforce. All three push AI agents, but ServiceNow's approach is a bit different.
The core difference is that AI is deployed by "role," not by "task". While typical AI agents handle individual tasks like "classify this ticket," Autonomous Workforce AI specialists take on entire roles like "L1 service desk agent." They're subject to the same access permissions, policies, and audit trails as human employees, performing the entire workflow of detecting problems, analyzing them, resolving them, and even updating the knowledge base.
Three AI specialist roles have been announced so far:
The first to launch, the L1 Service Desk AI Specialist, diagnoses issues by synthesizing infrastructure telemetry, observability tools, and security software data. ServiceNow says it doesn't follow scripts but understands context and reasons across systems. Nenshad Bardoliwalla (VP of AI Products) emphasized: "Our specialists are fundamentally different — they don't follow scripts, they understand context and reason across systems."
And then there's the AI Control Tower. It's a governance hub that monitors all AI agents across the enterprise — whether ServiceNow's own or third-party — from a single screen. You can see which AI is accessing which data, whether LLM providers are approved, whether there are prompt injection attacks, and how processing time, customer satisfaction, and escalation rates are trending. From a CIO's perspective, it directly solves the "we deployed AI but don't know what it's doing" problem.
EmployeeWorks — the result of the Moveworks merger
EmployeeWorks was born from combining the conversational AI and enterprise search technology of Moveworks, which ServiceNow acquired for $2.85 billion in December 2025. It's a "single front door for employees" — make natural language requests from Teams, Slack, browser, or mobile, and AI handles them across multiple systems. It targets roughly 200 million enterprise employees and is currently GA (generally available).
What changes?
Honestly, Salesforce Agentforce, Microsoft Copilot, and ServiceNow Autonomous Workforce might all look similar. But their approaches are quite different.
| Salesforce Agentforce | Microsoft Copilot Studio | ServiceNow Autonomous Workforce | |
|---|---|---|---|
| AI deployment unit | Task/topic-based | Per-app copilot | Role-based specialists |
| Primary domain | CRM / customer-facing | Office 365 productivity | ITSM / HRSD / SecOps back office |
| Governance | Einstein Trust Layer | Azure AI Content Safety | AI Control Tower (enterprise-wide AI command center) |
| Third-party agent mgmt | Limited | Within Azure ecosystem | ServiceNow + third-party AI unified monitoring |
| Workflow engine | Flow Builder | Power Automate | Now Platform (deterministic orchestration) |
| Employee frontend | Slack integration | Teams-centric | EmployeeWorks (Teams, Slack, web, mobile) |
ServiceNow's strength lies in back-office workflows. While Salesforce excels at customer-facing (front office) and Microsoft at document and email productivity, ServiceNow's core business is internal operations automation — IT service management, HR services, security operations. It also ranked #1 in the "AI agent building and management" category in the 2025 Gartner Critical Capabilities report.
And the numbers stand out. ServiceNow is already running Autonomous Workforce internally:
Handling 90%+ of employee IT requests with AI and resolving them 99% faster than humans are pretty aggressive numbers. Of course, this is ServiceNow's own environment, so whether the same results translate to customer environments remains to be seen. The L1 Service Desk AI Specialist is currently in Controlled Availability, with GA planned for Q2 2026.
Things to keep in mind
While AI Control Tower claims to manage third-party AI agents (Microsoft, Google, Salesforce, etc.), full visibility isn't guaranteed yet. And the 90%/99% internal metrics are based on standardized L1 cases like password resets and VPN tokens. Complex L2/L3 issues still need humans.
The essentials: how to get started
- Analyze your current IT service desk tickets
Identify the percentage of L1 tickets that are repetitive types (password resets, software installs, VPN issues). This becomes the ROI baseline for Autonomous Workforce adoption. - Review Zurich release upgrade
Autonomous Workforce is included in the Zurich release (March 2026). If your current ServiceNow version is Yokohama or earlier, start planning your upgrade. AI Control Tower is also available from Zurich onward. - Apply for L1 Service Desk AI Specialist
It's currently in Controlled Availability. Request early access through your ServiceNow AE. GA is planned for Q2 2026. - Pilot EmployeeWorks
Moveworks-based EmployeeWorks is already GA. Connect it to existing channels like Teams and Slack to create a single front door where employees can handle IT and HR requests in natural language. - Set up governance with AI Control Tower
Before scaling up AI agents, define in AI Control Tower who has access to what data, which LLM models are approved, and what the escalation policies are.
