What Happens to L1 Support When the AI Specialist Shows Up

At Knowledge 2026 in May, ServiceNow rolled out what it's calling the Autonomous Workforce, a set of AI Specialists built to run inside real business functions rather than sit as a chat window bolted onto a workflow. The one aimed squarely at entry-level ServiceNow careers is the L1 IT Service Desk AI Specialist, and it's worth taking seriously rather than dismissing as vendor hype, because the company backed the launch with named customers and specific numbers instead of just adjectives.

The City of Raleigh's CIO, Mark Wittenburg, told ServiceNow that its virtual agent, nicknamed Ral-E, now deflects 98 percent of requests before they reach a human tech, saving what he described as the equivalent of a full month of staff time. DocuSign's global VP of IT, Saran Mandair, said the company's goal is to autonomously handle 90 percent of all tickets so human agents can focus on harder problems. Honeywell's SVP and chief digital technology officer, Sheila Jordan, said its AI assistant, called Red, eliminated the majority of routine service desk conversations. ServiceNow's own internal help desk, which the company uses as a testbed for its own products, reports the L1 AI Specialist resolving assigned cases 99 percent faster than a human agent working the same case. In a separate interview with CIO.com, SVP of central product management John Aisien said the tool is currently handling more than 90 percent of employee requests inside ServiceNow itself, almost twice as fast as a human agent.

Those are the vendor's own reported figures, worth reading as best-case numbers rather than an independent audit. An analyst quoted in that same CIO.com piece flagged the obvious caveat: getting anywhere close to those results at another company requires real groundwork first, cleaned-up knowledge bases, mapped workflows, and historical incident data good enough to train on, and most IT shops aren't there yet. Still, the direction isn't really in question. Simple, repetitive L1 requests, password resets, access provisioning, my VPN won't connect, are exactly the category of work agentic AI handles well, and ServiceNow is far from the only vendor building toward this.

This isn't only a ServiceNow story

Gartner surveyed 110 chief human resources officers in mid-2026 and found that 22 percent said at least one business leader in their organization had already stopped hiring for entry-level roles because of AI automation, even though the same survey found only about a fifth of companies had gotten what Gartner called significant or transformational value out of their AI rollouts so far. Gartner's Kaelyn Lowmaster put the risk plainly: organizations that respond by cutting their early-career pipeline altogether are trading a short-term cost save for a longer-term talent problem. That's a general labor-market signal, not an IT-specific or ServiceNow-specific one, but it lines up with what the L1 service desk numbers suggest. The easiest rung of the entry-level ladder is the one getting automated fastest, in IT and elsewhere.

So if you're considering L1 support as your way into a ServiceNow career, here's the practical read: the door hasn't closed, but the version of L1 that pays you to answer the same handful of ticket types for a year or two while you slowly absorb platform knowledge is getting shorter and less reliable as a training ground. The organizations rolling out an AI Specialist still need people. They need people who can configure the thing, tune its knowledge base, handle the tickets it correctly escalates because they're genuinely not routine, and audit its decisions when something goes sideways. That's a different skill set than answering the phone and following a runbook, and it's a more durable one.

Practically, that argues for spending less energy chasing entry-level help desk postings as the whole plan and more time building toward the skills that sit just above the automation layer. Learn the CMDB well enough to understand why an AI agent's routing decision was right or wrong, not just how to update a record in it. Get comfortable with the tables that actually drive incident, problem, and change management, because someone still has to configure and maintain the logic an AI Specialist runs on top of. Pick up enough scripting, business rules, client scripts, and basic Flow Designer or Now Assist configuration to be the person who builds and adjusts the automation rather than the person the automation was designed to replace. A CSA certification paired with a portfolio that shows you can actually configure and reason about the platform's data model puts you in a materially different conversation with a hiring manager than wanting to start in a help desk seat and work your way up, because a lot of those seats are exactly the ones shrinking first.

None of this means L1 support work vanishes overnight, and it doesn't mean the field is closing to newcomers. What it means is that the old slow-ramp path, start at L1, absorb the platform by osmosis over a couple of years, is becoming a less dependable way in. The people who get in cleanly from here are the ones who can show up already understanding the system well enough to work alongside an AI Specialist instead of competing with one for the same queue.

The Platform Path is being built around exactly this kind of grounded, honest read on where the real entry points into ServiceNow work actually are right now. If you want in when doors open, the early-access waitlist is right here on this site.