Every year, ServiceNow Knowledge generates more announcements than any organization can act on. Our job (and yours) is knowing which ones actually matter right now, which ones are worth watching, and which ones can wait. We spent our time in Las Vegas making that call. Here’s what we came back with.

The Platform Is Becoming an Operating System for Work

ServiceNow is no longer positioning itself as a workflow tool that added AI. It’s positioning AI as the foundational layer on which everything else runs. That might sound like marketing language, but it’s a meaningful architectural shift, and it has real implications for how you should be thinking about your platform investment.

The most concrete expression of that shift is Otto, ServiceNow’s unified agentic execution framework. Otto connects AI agents to workflow elements across the entire platform, enabling an agent to reason, act, and hand off work across IT, HR, security, and operations without human intervention at each step. For leaders, the practical implication is this: the question has shifted from “which process can we automate?” to “how do we govern work that AI is executing on our behalf, and what happens when it gets it wrong?”

Most organizations have yet to answer that question. Those who do will scale, and those who don’t will find themselves managing incidents instead of outcomes.

Governance is Now a Security Problem, Whether You’re Ready or Not

If Otto is the engine, the AI Control Tower is the dashboard and the guardrails. With real-time visibility into agent activity, audit trails, kill switches, and permission guardrails, it’s the accountability layer that makes autonomous AI deployable in a regulated, risk-aware enterprise.

Governing AI agents isn’t a platform administration challenge with a security dimension. It’s a security challenge with a platform dimension. This distinction matters, because it changes who needs to be in the room and what “done” looks like.

As AI agents proliferate, they create non-human identities (e.g., service accounts, machine identities, agent credentials) that can outnumber human users by ratios of 45:1 or higher. These identities are largely invisible to traditional security tooling. They bypass conventional access controls, carry permissions that were never designed for autonomous actors, and too frequently have no defined owner. Organizations that deploy AI agents without solving this problem first might look like they’re moving fast, but in actuality, they’re creating security debt that they’ll spend years unwinding.

Two partner integrations give the Control Tower meaningful teeth here. Veza brings non-human identity governance into the picture, ensuring AI agents operate under least-privilege access with continuous monitoring. Armis extends asset intelligence into OT and IoT environments, closing the visibility gap between IT security and physical infrastructure that has existed for years but now poses a threat when AI agents can interact across both.

This is precisely why AHEAD approaches AI Control Tower as a joint motion between our ServiceNow and Governance & Cybersecurity practices — not two separate conversations that happen to touch the same platform.

A More Welcoming Front Door for Employees

On the employee side, EmployeeWorks and Employee Slate were among the more practical announcements at the conference.

ServiceNow has struggled with the same self-service problem that many enterprise platforms have faced for years: employees will not use a portal consistently if the experience feels like work. Most self-service environments have asked users to adapt to the system instead of the other way around.

Employee Slate is a meaningful improvement because it brings together role-based context, pending work, and relevant services in a more usable front door. Paired with Moveworks, it gives employees a more natural way to ask for help and get something done.

That matters, as self-service adoption rarely fails because employees are willfully resistant. More often, these failures can be traced back to experiences that are fragmented, generic, or easy to avoid. This release does not solve that problem on its own, but it moves in the right direction.

Autonomous Operations Are Only as Strong as the Signals Behind Them

The push toward autonomous IT operations was a dominant theme at K26. The vision, where the platform detects, diagnoses, and remediates issues without human triage, is real and closer than many IT leaders expect. However, it has a hard dependency: you need telemetry you can trust before you can act on it autonomously.

Dynatrace played a prominent role in that story, and for good reason. Event data, topology context, and anomaly detection are what allow ServiceNow workflows to act with confidence. Without that signal quality, autonomous operations become a faster way to reach the wrong conclusion.

This is one reason AHEAD’s Intelligent Operations practice matters in this conversation. It connects ServiceNow, Dynatrace, and Tanium in a way that makes the data, the workflow, and the response model work together.

The Autonomous Workforce: Further Away Than Most Think

K26 introduced role-specific AI agents for particular functions, with unique context, vocabulary, and decision criteria built in. It’s a more sophisticated framing than the general-purpose agent conversation that dominated last year, and it’s the right direction.

Our view is that the ‘autonomous workforce’ discussion has some time before becoming a primary execution challenge for most enterprises. Right now, it’s a design and governance challenge. The organizations we see moving fastest aren’t the ones who’ve deployed the most agents — they’re the ones who’ve done the hard, unglamorous work of defining what human judgment is actually for, redesigning workflows around AI participation, and building the change management muscle to bring their workforce along.

The K26 sessions on this topic were more grounded than we expected, and we appreciated it. But the gap between the vision on stage and the operational reality in most enterprises is still significant. If you’re a leader trying to set expectations internally, the right message isn’t “we’re behind.” It’s “the organizations winning here started with foundation, not features.” So should you.

The Uncomfortable Truth: Most Enterprises Aren’t Ready for What ServiceNow is Becoming

This was the most repeated message across K26, and it’s the conversation we have most often with clients: AI priorities are rising faster than platform readiness, and the mismatch is creating real risk.

This is a familiar pattern. Leaders are under pressure to show AI progress, so pilots get launched and the results come back uneven. The culprit is rarely the AI capability itself, but the environment it’s operating in: fragmented CMDBs, over-customized instances, inconsistent data quality, unclear ownership, and underused capabilities that should be doing the heavy lifting already. These aren’t new problems, but they’re newly consequential because they’re now blocking AI value, rather than just reporting accuracy or operational efficiency.

We’ll say this plainly: if your ServiceNow environment wasn’t built to the governance and data quality standards the platform now requires, layering agentic AI on top of it won’t accelerate your outcomes. It will accelerate your problems.

The good news is that this is a sequencing challenge, not a fundamental one. AHEAD’s Data Foundations practice addresses the data layer directly. But data readiness is only part of the picture.

The broader question we help leaders answer is where exactly is your platform underperforming, and what has to be true before your AI investments can scale? That’s the core of our NOW Platform Optimization & AI Readiness engagement, a combined advisory motion that’s deliberately not just a platform assessment and not just an AI strategy workshop. It surfaces where ServiceNow value is leaking today, baselines your AI readiness across strategy, data, governance, process, talent, and operating model, and produces a ranked set of use cases and pilot candidates with clear prerequisites. The output is a short, leadership-ready decision framework — not another slide deck of observations, but a concrete answer to the question: what do we do first, and why?

Final Thoughts

Knowledge 2026 confirmed something we’ve believed for a while: the ServiceNow platform is evolving faster than most organizations’ ability to absorb it. That’s not a criticism of the platform — it’s a planning reality. Speed will not determine which organizations will get the most value from the platform over the next 18 months. Enterprises poised for success are the ones strengthening data foundations, building governance in early, and prioritizing use cases their environments can support.

If there is uncertainty about where your organization stands today, embrace it. Get in touch with AHEAD to figure out what is ready, what is not, and what should come next.