News
White Paper
Why AI Rarely Improves How Work Actually Gets Done
ServiceNow thinks it’s cracked the code on autonomous agents, bringing it one step closer to its vision of being the center of all AI operations — or, as it likes to put it, the “AI control tower for business reinvention.” Autonomous Workforce is the company’s new solution which pushes AI agents beyond narrow task-based use cases to handle role-based work from end-to-end.
The company also launched ServiceNow EmployeeWorks, the result of its Moveworks acquisition finalized in December, to help flesh out this vision. EmployeeWorks brings in the conversational AI interface and enterprise search of Moveworks to give employees a single front door to solve problems. EmployeeWorks essentially acts as a middle man, ingesting employee requests, issues and queries, to then assign tasks to the relevant agent for each task and coordinate handoffs. The idea is EmployeeWorks relieves employees of the need to figure out which agent performs which task.
A Role-Based Autonomous Workforce
The Autonomous Workforce launch included the first out-of-the-box specialist of what the company promises will be many, the L1 Help Desk Specialist.
Nenshad Bardoliwalla, Group VP, AI Products, called the L1 Help Desk Specialist the company’s “proof of concept” for an entire workforce that will assist teams across every department. The company cited employee services, security operations, customer service, finance and legal as future specialists in the works.
The L1 Help Desk Specialist is designed specifically to handle workplace technology incidents from diagnosis to determine the resolution to taking action and then closing the ticket. It updates the employee who reported the incident as well as logging the activity in the knowledge base — all without human intervention. These specialists are already being used internally at ServiceNow and in beta testing in the City of Raleigh.
What sets these autonomous workers apart from bots or single-use AI agents is they operate within a clearly defined role. The specialists are non-deterministic, meaning they don't follow a prescribed path to reach resolution, and instead adapt to the best method for a given case to resolve an issue without human intervention.
One of the guarantees underlying this — and one that is table stakes for any IT director to set an autonomous agent loose in their stacks — is ServiceNow's governance layer. What this means in practical terms is any AI specialist observes existing rules and restrictions, but is also monitored by the AI Control Tower for model drift to ensure it plays by the rules.
ServiceNow also designed the specialists to do something that is currently a rarity among LLMs: acknowledge when it can't do something. In cases where the incident surpasses the available information, the specialist takes the resolution as far as it can and then hands the task off with background documentation to the appropriate human L2 or L3 specialist for resolution.
A New 'Front Door': ServiceNow EmployeeWorks
ServiceNow acted quickly to incorporate its Moveworks buy into the stack. The result, EmployeeWorks, is the AI chat interface we're now growing familiar with but with a key difference from most: EmployeeWorks doesn't just summarize, it oversees the completion of task. Employees converse in natural language with EmployeeWorks from any application they're working in — Slack, Microsoft Teams, etc — and it then assigns actions to the appropriate autonomous workflows behind the scenes.
Moveworks CEO and founder Bhavin Shah joined ServiceNow as SVP GM of Moveworks when the acquisition closed in December. He framed EmployeeWorks as an antidote to the disconnected experiences and fragmented AI tools that now exist as businesses raced to plug on individual AI apps into existing SaaS platforms. EmployeeWorks doesn't just connect across all business systems, it can also access and search web data to pull in real-time information in response to employee queries.
Moveworks will continue to be offered as a standalone option under the ServiceNow umbrella, meaning no interruption for existing customers and less of a commitment for potential customers.
What ServiceNow Sees as Its Differentiator
Where some vendors may be shedding past services and capabilities to adapt to the current AI zeitgeist, ServiceNow sees its background in IT service management as a strength to build on. Because with ITSM, you're creating a connective tissue across all departments — HR, customer service, legal, procurement, etc — in order to see tickets through to completion.
The connective tissue built up over time from these workflows — the CMDB or configuration management database — provides the information for AI to understand how to resolve issues. Bardoliwalla said the two decades of structured data built up in the CMDB is what helps ServiceNow overcome the poor documentation which thwarts other AI initiatives.
"This isn't a system that reads your Word docs and hopes for the best. It's built on top of a live CMDB, active workflows, policy engines, approval chains and real transaction history. The "documentation" for ServiceNow isn't a PDF someone wrote in 2019 and forgot to update. It's the actual system of record for how your enterprise operates, updated in real time, every time a ticket closes, a workflow runs or a policy is updated," Bardoliwalla said.
When you add ServiceNow's well-established governance and integrated risk management layer on top of the data and knowledge built up from workflows, it gives agents both the information and patterns to learn from and the guardrails to keep them in line.
ServiceNow's AI Control Tower — launched at Knowledge 2025 — extends that governance layer over its AI offerings. AI Control Tower is vendor neutral, meaning the AI oversight, management and security protections extend to any AI in use, be it ServiceNow's or a third-party agent, model or workflow. The company made the intentional choice to remain an open platform, meaning partners can build agents on top of it.
Webinar
Mar
26
Do Learning Programs Really Work? How to Turn Education Into Engagement In Healthcare
See how leaders are using learning programs to build trust with healthcare professionals and create measurable engagement.
Register
Webinar
On demand
Fuel Your Front Line: Workforce Trends Shaping 2026
Get expert insight into the tools and tactics top frontline teams will need in 2026.
Watch Now
Webinar
On demand
Fix the Disconnect Between Projects and Profit
Your margins are hiding in disconnected data. Let's find them.
Watch Now
Webinar
On demand
AI in Customer Service: Faster Resolutions, Happier Customers
Don’t let rising demand burn out your team. See how to build a smarter, more resilient support org.
Watch Now
What Next?
ServiceNow is in an odd, though not necessarily bad, place among enterprise platforms angling to be the AI platform of choice. Its closest competitor, Microsoft, has the advantage of the broad reach of its collaboration and productivity tools, making it the center of daily work in a majority of enterprises. While criticisms of Copilot have not been lacking, if Microsoft improves it to a place where it is "good enough," will ServiceNow be able to convince people to move their daily operations into a new platform?
Salesforce is a recently formed rivalry, but one that both Bill McDermott and Marc Benioff appear to relish. With Salesforce's move into the back office with its ITSM offering and ServiceNow's move into CRM, the rivals are threatening to meet somewhere in the middle. Benioff admitted to a Gartner IT audience last fall that Agentforce adoption has been slowed in part due to confusion around pricing, and complaints around lost conversational context have also been reported.
Today's Autonomous Workforce launch and EmployeeWorks are ServiceNow's most visible answers to those competitive questions so far. But they're one piece of a much larger bet — the company has spent billions on acquisitions in the past year alone, each targeting a different layer of the enterprise AI stack. The strategy is coherent. EmployeeWorks shows ServiceNow has the ability to quickly integrate an acquisition. Next comes the harder part: integrating the rest of the stack and proving the AI Specialists can scale beyond a handful of beta testers into a broader Autonomous Workforce in general availability.
Have a tip to share with our editorial team? Drop us a line: