We have been doing a lot of work in the ServiceNow space lately, and so many of the conversations we are having jump straight into the same topic: AI in production, with actual workflows, risk, and responsibility.
And the pattern holds across platform owners, workflow leads, and automation engineers alike. Once AI starts touching real business processes, the questions get serious fast.
For a couple of years, generative AI on the ServiceNow platform meant something fairly contained: a case got summarised, a next step got suggested, and a person still acted on it. That was the era of Now Assist doing useful, bounded work with a human firmly in the loop, often shortened to HITL in governance discussions.
What has changed is autonomy. Agentic AI does not just suggest an action, it reasons through a request, plans the steps, and executes them. A generative assistant tells you a password needs resetting. An agent resets it and closes the ticket itself. That shift, from assistance to execution, is why the roles built around ServiceNow are changing shape.
It also explains why ServiceNow’s own product direction has moved so heavily toward governance infrastructure. At its Knowledge 2026 conference, the company positioned its AI Control Tower as a way to discover, govern, observe, secure, and measure every AI agent and model running across an enterprise, regardless of where that agent originated. Alongside it came Otto, a unified AI brand bringing Now Assist, the company’s Moveworks acquisition, and its conversational AI Experience layer under one entry point. The message was consistent: intelligence is not the constraint anymore. Governance is.
How This Is Reshaping the Roles Companies Hire For
Platform Owners, who used to focus primarily on stability and uptime, are increasingly making governance calls: what can be automated, what data an agent is allowed to touch, and where a human still needs to stay in the loop before an action executes. Tools like AI Control Tower and Now Assist Guardian exist precisely because this decision now sits with a named accountable person, not a policy document nobody reads.
Workflow Designers are starting to think like reliability engineers. Confidence scoring, fallback logic, and rollback paths are being built into flows that never needed them when a human was the one making every judgment call. If an agent is not confident enough in a decision, something has to catch that gracefully, and that something has to be designed in advance.
Developers are being pulled into cost strategy in a way that did not exist eighteen months ago. Most generative AI actions on the platform consume metered Assist units, and every custom skill, every fallback trigger, every summarisation call adds up against that pool, including in sub production and testing environments. Optimising token usage, choosing the right model size for the task, and rewriting prompts for efficiency are no longer nice to have skills. They are now directly tied to a line item on someone’s budget.
Why the Data Underneath Still Decides Everything
None of this works without clean data underneath it. A CMDB (Configuration Management Database) that is genuinely aligned to the Common Service Data Model, or CSDM, is what makes impact analysis and agentic action reliable rather than guesswork. The pattern practitioners keep repeating is blunt but accurate: AI surfaces whatever is already there. If the knowledge base, the CMDB, and the historical ticket data are inconsistent, the AI simply produces inconsistent output faster than a person would have.
This is the reality we are hearing directly from the people doing the work, not from a roadmap slide. AI in production is resetting what platform owners, workflow designers, and developers are actually accountable for inside ServiceNow, and the roles companies are hiring for now reflect that shift.
If you are living this shift day to day, whether in ITSM, ITOM, HRSD (HR Service Delivery), CSM (Customer Service Management), or SecOps, we would genuinely like to hear how it is playing out on your side. As a ServiceNow recruitment agency, Brabers is always glad to compare notes.
