The AI job market has stopped being one story. It’s two stories running in parallel, in the same economy, about the same technology — and they barely overlap.
In one, experienced professionals are watching their wages climb. The Dallas Fed reports that workers in AI-exposed sectors like computer systems design have seen wage growth of 16.7% since fall 2022, more than double the 7.5% national average. Senior engineers, seasoned lawyers, and veteran marketing specialists are using AI tools to work faster, take on more complex projects, and command higher pay.
In the other, the class of 2026 is walking into a wall. Goldman Sachs calculates that AI is erasing roughly 16,000 net U.S. jobs per month. BlackRock CEO Larry Fink calls it a “crisis” for new graduates. The share of unemployed Americans who are new workforce entrants hit a 37-year high in 2025.
Same technology. Opposite outcomes. And the gap is widening.
The Numbers
The Q1 2026 data is now complete, and it’s brutal. 78,557 tech workers lost their jobs in the first three months of the year. Of those, 37,638 — nearly 48% — were attributed to AI and automation. In March alone, AI was the number-one cited reason for job cuts for the first time ever.
But who’s actually getting cut? That’s where the data splits.
Dallas Fed researchers found that workers ages 22 to 25 in AI-exposed occupations experienced a 13% decline in employment since 2022. For software developers in that age bracket, the drop is closer to 20%. Meanwhile, employment totals for older workers in the same occupations haven’t declined at all.
Goldman Sachs breaks it down further: AI substitution eliminates roughly 25,000 jobs per month, while AI augmentation creates about 9,000. The net negative falls almost entirely on entry-level positions — data entry, customer service, legal support, billing — the exact roles where Gen Z workers are concentrated.
Why Experience Is the Dividing Line
The Dallas Fed’s research explains the mechanism. AI is good at automating “codifiable” tasks — things you can learn from textbooks, online courses, or training manuals. That’s exactly what entry-level jobs consist of. A junior developer writes code that fits established patterns. A new paralegal does document review. A first-year analyst builds spreadsheet models.
AI does all of that now.
What AI can’t replicate is tacit knowledge — the judgment and intuition built through years of hands-on work. Knowing which client is bluffing, when a project is heading off the rails before the metrics show it, how to navigate internal politics to get a decision made. That’s the stuff that makes experienced workers more productive with AI, not replaceable by it.
The result is a widening wage gap. A one standard-deviation increase in AI exposure widens the entry-level-to-experienced wage gap by roughly 3.3 percentage points. In plain terms: the more AI touches your industry, the more it pays to already have ten years of experience — and the less it pays to be starting out.
The Responses
Different players are reacting to this split in very different ways.
Oracle cut 30,000 employees to free up $8-10 billion for AI data center construction. Termination emails arrived at 6 a.m. with no prior warning from HR. India lost 12,000 positions. This is the most straightforward version of the trade: humans out, infrastructure in.
IBM went the other direction, tripling its entry-level hiring. But here’s the catch — they rewrote the job descriptions. Junior developers now spend less time on routine coding and more time working directly with customers. IBM’s CHRO Nickle LaMoreaux argues that companies slashing entry-level roles now will face a talent pipeline crisis in five years when they need mid-level managers and can only poach from competitors. She might be right.
Gen Z itself is adapting through what researchers call “polyemployment” — working multiple part-time jobs simultaneously, a trend that’s hit its highest level in over a decade. Some are doing it by choice, using AI tools to juggle multiple gigs efficiently. Others are doing it because full-time entry-level positions no longer exist in their field. The data from Deputy’s “Big Shift 2026” report shows Gen Z accounts for two-thirds of workers holding multiple jobs in the UK.
And then there’s the 44% of Gen Z workers who told researchers they’re intentionally sabotaging their company’s AI rollout. Faced with a technology that threatens their livelihood while enriching their senior colleagues, some have decided passive resistance is the rational response.
The Pipeline Problem
Here’s what none of the executives firing entry-level workers seem to be thinking about: where do experienced workers come from?
The career ladder has rungs for a reason. You hire a junior developer at 22, they become competent at 25, skilled at 28, senior at 32, and a lead at 35. That pipeline takes a decade. If you stop feeding the bottom of it, you don’t feel the pain for five to seven years — then it hits all at once.
IBM’s LaMoreaux is one of the few corporate leaders saying this out loud. “The companies three to five years from now that are going to be the most successful are those companies that doubled down on entry-level hiring in this environment.” BlackRock’s Fink is investing $100 million in skilled-trade programs, betting that the white-collar entry-level collapse will push more workers toward data center construction, electrical work, and infrastructure — fields that AI can’t automate and that AI’s growth actually requires.
Goldman Sachs offers one piece of good news from 40 years of displacement data: younger, college-educated, urban workers who lose jobs to technology typically experience earnings losses roughly half as severe as older displaced workers over the following decade. Gen Z may be getting hit first, but historically, their cohort recovers faster.
Of course, no previous generation faced a technology that got meaningfully more capable every three months.
What This Means
The AI job market isn’t a single wave washing over everyone equally. It’s a sorting machine, and experience is the sorting criterion.
If you have ten-plus years in your field, AI is probably making you more productive and more valuable. Your wage trajectory is likely positive. The Challenger report found hiring plans surged 157% in March. AI engineer salaries average $206,000, with MLOps specialists commanding $165,000+ and top research positions exceeding $489,000 in total comp.
If you’re under 25 and entering the workforce, you’re facing conditions that haven’t been this difficult since the late 1980s. The traditional playbook — get a degree, land an entry-level position, work your way up — is cracking under the weight of a technology that can do junior work for free.
What You Can Do
If you’re starting out: Stop competing with AI on its terms. A chatbot can write boilerplate code, draft standard contracts, and build basic models. It can’t sit in a room with a difficult client and de-escalate a situation. Optimize for roles that put you in front of humans, not screens. IBM’s model — rewriting junior roles to emphasize customer interaction — is the template to look for.
If you’re mid-career: You’re in the sweet spot, for now. Use this window to build the tacit knowledge that makes you AI-resistant. Learn the tools well enough to 2x your output, but don’t assume the advantage is permanent. AI capabilities are moving upward through the experience ladder, not staying at the bottom.
If you’re hiring: IBM is making a bet that most of your competitors aren’t. The entry-level talent pool is the most available it’s been in years, and the cost of ignoring it compounds annually. Companies that build their pipeline now will have a structural advantage when everyone else is fighting over the same pool of senior talent in 2030.
The next data point is the April Challenger report, due in early May. March was already a record. Whether April breaks it again will tell us if this divergence is stabilizing or accelerating.