The IT Career Ladder Is Missing Its Bottom Rung
- CyberSpeak Labs
- 1 hour ago
- 3 min read
By Stephen Engler, Owner of Ctrl-Alt-GenX and cybersecurity speaker (@PHtheAdmin)
I started in IT the way almost everyone I know started in IT: answering the help desk phone.
Password resets, "the printer's down again," the occasional actual emergency buried in forty
routine tickets. It wasn't glamorous. But it's where I learned to triage under pressure, talk to a scared end user, and build the instincts that eventually let me do harder things. Every senior
engineer I've ever worked with has some version of that same story.
That story is getting a lot harder to tell in 2026.
AI agents are now handling the exact tickets that used to be someone's first job. Across
enterprise support programs, the median tier-1 deflection rate sits at 41.2%, with the top quartile clearing nearly 59%. Password resets and refund requests, the bread and butter of an
entry-level queue, deflect at 70% or higher. McKinsey puts the cost difference at $0.62 for an AI resolution versus $7.40 for a human one. From a spreadsheet, that's not even a close call.
From a career-pipeline standpoint, it's a problem nobody budgeted for.
Here's the part that should worry all of us more than the ticket stats. Stanford's Digital
Economy Lab dug into ADP payroll data and found that employment for workers age 22 to 25
in the most AI-exposed occupations, software developers, customer service reps, IT support,
is shrinking at over 4% a year. Meanwhile, employment for that same age group in less
AI-exposed jobs is growing 2% a year. The researchers drew a distinction that stuck with me: jobs built on codified knowledge, stuff you can look up or automate, are the ones disappearing for young workers. Jobs that depend on tacit knowledge, the kind you only get through mentorship, repetition, and being in the room when something breaks, are holding steady or growing. Tier-1 help desk work was tacit-knowledge training disguised as codified-knowledge grunt work. We're automating away the disguise and losing the training with it.
It's not just fewer tickets to answer. PwC's 2026 AI Jobs Barometer found something they're
calling "seniorization." Companies aren't killing entry-level job titles outright, they're quietly stuffing them with senior-level requirements instead. In the most AI-exposed roles, 52% of the
new skills showing up in entry-level postings are things that used to be reserved for experienced hires: judgment calls, stakeholder management, strategic thinking. In the least-exposed roles, that number is 7%. Nearly half of hiring managers PwC surveyed said they'd rather invest in AI tools than hire and train a recent grad. And a Resume.org survey found 21% of companies have already frozen entry-level hiring specifically because of AI, with 36% saying they'll stop entirely by the end of this year.
CompTIA's State of the Tech Workforce report shows the twist in all this. Overall tech employment is still growing, projected up 1.9% this year, adding about 185,000 jobs. The field
isn't shrinking. But entry-level postings are down while the applicant pool has exploded, to the point where entry-level developers report sending out 200-plus applications without a single response. Growth at the top, a bottleneck at the bottom. That's not a healthy pipeline, that's an hourglass.
This is exactly what I mean when I talk about human infrastructure. A help desk was never
just a cost center to be optimized down to zero. It was the training ground, the place where a
22-year-old learned how to stay calm on the phone with someone who's furious, how to spot
the ticket that's actually a security incident wearing a "my email won't open" costume, how to build the judgment that no AI model has, because judgment isn't codified knowledge. You
don't get it from a manual. You get it from a thousand small reps, ideally with someone senior enough to occasionally clean up your mistakes without letting the building burn down.
If you automate that rung away without building something to replace it, you're not eliminating
a cost. You're eliminating your own future senior staff. The incident commanders and infrastructure leads you'll need in ten years were supposed to come from somewhere, and
"somewhere" used to be a help desk queue.
So here's the question I'd put to any IT or security leader reading this: if AI is taking your tier-1 tickets, and it probably should, what are you deliberately building to replace the training that used to happen there? Shadowing rotations, structured mentorship, simulated incidents, something? Or are you just hoping the next generation of senior engineers shows up fully formed, because you stopped growing your own?
Because right now, most companies haven't answered that question. They've just stopped
hiring for the job that used to ask it.
Stephen Engler
Owner, Ctrl-Alt-GenX | Cybersecurity Speaker @PHtheAdmin
Comments