From Advisory AI to Execution
ServiceNow set the tone for its Autonomous Workforce push with a line that doubles as a strategy: advisory AI has run its course, and enterprises need AI that senses, decides and securely acts within organizational guardrails. That framing, from president and chief product officer Amit Zavery, lands harder in June 2026 as the company moves additional IT AI specialists into general availability. These are not chatbots bolted onto the platform. They are designed to complete end to end processes alongside humans, autonomously resolving cases, containing threats and managing incidents across the IT estate.
The June milestone matters because it extends autonomy past the obvious entry point. The L1 IT Service Desk specialist has been available and, ServiceNow claims, resolves cases 99 percent faster than human agents. Now the company is shipping specialists for AIOps, site reliability engineering, asset lifecycle and portfolio planning, while putting security and risk specialists into preview this month ahead of general availability in September. The pattern is deliberate: prove value on high volume, low judgment work first, then push agents into territory that touches reliability and risk.
The Numbers Customers Are Reporting
The customer evidence is what separates this from the usual agentic marketing. Sheila Jordan, senior vice president and chief digital technology officer at Honeywell, said the company's AI assistant eliminated the majority of service desk conversations, saving time for both employees and the IT organization. Saran Mandair, global vice president of IT at Docusign, set the bar even higher, describing a goal of autonomously handling 90 percent of all tickets so human agents can concentrate on the most critical work. Mark Wittenburg, CIO of the City of Raleigh, reported saving the equivalent of a full month of time through documentation and automation alone.
The platform scale numbers reinforce the point. ServiceNow says its Autonomous CRM already resolves more than 100 million cases monthly, orchestrates over 16 million orders and configures more than 7 million quotes, while its employee portal generates over 40 million cases a year across 23 million monthly users, with 91 percent resolved without reassignment. Whatever skepticism we hold toward vendor metrics, throughput at that magnitude is a credible signal that ServiceNow is operating these specialists in production, not in a lab. For CIOs evaluating where to place bets, demonstrated volume beats demo polish.
The Platform War Over Who Acts
This release is best read as a move in a larger contest. SAP is shipping its Joule agents, Oracle has layered agentic capability through Fusion, Salesforce is extending Agentforce, and Microsoft is pushing Copilot deeper into business workflows. Every one of these vendors is racing to own the layer that does not just hold enterprise data but acts on it. ServiceNow's advantage is that it already sits as a system of action across IT, HR and customer workflows, which is a more natural home for execution agents than a system built primarily to record transactions.
The risk, predictably, is governance. Pushing autonomous specialists into security operations and risk screening, as ServiceNow now plans, means agents will make or recommend decisions with real blast radius. The same week's IBM research showed that organizations relying on manual governance averaged dozens of agent incidents last year. ServiceNow has answered with its AI Control Tower and an autonomous security and risk product governing agent identities and permissions, but CIOs should treat those controls as the thing to evaluate first, before they evaluate any specialist's speed.
Our Read for Enterprise Buyers
For organizations already standardized on ServiceNow, the June availability is a low friction opportunity to test execution agents on processes where the cost of error is contained, such as L1 tickets and asset lifecycle tasks. The harder strategic call is whether to let a single platform vendor own execution across IT, CRM and security, or to keep that capability distributed. Concentration buys integration and a single governance pane; it also deepens lock in at exactly the moment switching costs are rising across every agentic platform.
Our view is that the buyers who win this cycle will be the ones who insist on measurable outcomes and auditable controls before they expand scope. A specialist that resolves tickets 99 percent faster is worth little if it cannot be observed, constrained and rolled back when it errs. ServiceNow has built a credible execution story and the production volume to back it. The discipline now sits with the enterprises deploying it, who must resist the temptation to mistake autonomy for a finished governance model.
