ServiceNow, Inc. (NYSE: NOW) has expanded its EmployeeWorks offering with AI assistants, employee-built routines, developer tooling and an API intended to put governed AI work inside the applications employees already use. The ServiceNow EmployeeWorks AI assistants update targets a familiar enterprise constraint: staff closest to a business process often cannot build an AI workflow because of IT backlogs, developer capacity, fragmented tools and governance requirements.
The company said the expanded platform is in controlled availability. It is built on what ServiceNow calls a third-generation employee-centric AI harness, combining context, reasoning, memory and tool interoperability. ServiceNow positions that foundation as a way to let employees tailor AI behavior, content, tools and access while administrators define guardrails.
Employee-Built Routines Move Repetitive Work Into Governed Workflows
For repeatable tasks, EmployeeWorks now includes Productivity Routines, which employees can build and deploy to run proactively in the background. ServiceNow said users describe the work they want automated, select the tools they need and deploy the routine without needing to understand backend systems, API integrations or data models.
That distinction matters for enterprise teams. A no-code path may reduce the handoff between the people who understand a process and the people who configure it, but it does not remove the need for access controls and operational review. ServiceNow says the platform uses administrator-defined guardrails; its source does not establish how each organization should configure those controls or which employee-created routines are appropriate for production.
ServiceNow also introduced Productivity Assistants for more contextual work. The company said prebuilt and user-configured assistants can support long-running tasks and work across systems and agents. Initial HR Service Delivery uses include assistants for talent insights, HR business-partner work and manager coaching, while two foundational EmployeeWorks assistants handle research and calendaring.
ToolSmith Gives Developers a Faster Integration Path
Employee-facing tools still need dependable connections to enterprise systems. ServiceNow’s new ToolSmith developer environment is designed to build automations for EmployeeWorks assistants from natural-language instructions. According to the company, ToolSmith is trained on more than 75,000 APIs, develops a plan with the developer, implements integrations using its best practices and can retrieve relevant logs and propose a fix when something fails.
Those capabilities could make ToolSmith relevant to IT teams balancing demand for custom assistants against the work of building, testing and maintaining integrations. ServiceNow also announced a Conversations API that can bring EmployeeWorks capabilities into custom internal applications, kiosks and other AI applications. The company said this approach does not require organizations to rearchitect those channels, though buyers will need to validate that claim against their own integrations and identity model.
Why Governed Employee AI Changes the Operating Model
ServiceNow has made the launch a platform play rather than a standalone assistant release: employees can create and use AI capabilities, developers can extend them, and the resulting tools can surface in existing channels. The company cites a ServiceNow customer and partner that reported lower request-submission time and a higher completion rate, but it did not provide independent performance data for the new functions.
The larger shift is from legacy employee portals and one-off automation requests toward an AI layer that can connect work, approved tools and enterprise context. That does not eliminate the implementation burden of data access, guardrails and workflow ownership. It does, however, give ServiceNow a way to compete on the integration of employee AI with governed workflows instead of treating assistants as isolated chat interfaces.