AI Cut 13% of Entry-Level Hiring and Marketing Isn't Budgeting for 2031
Entry-level US workers ages 22-25 saw employment fall 13% in the most AI-exposed jobs from late 2022 through mid-2025, according to Stanford's Digital Economy Lab. Boston Consulting Group now estimates 90% of marketing manager tasks face some level of AI disruption. Teams cutting junior hires today are quietly pricing in a senior-talent shortage for 2030-2031 with no plan to fund it.
The 13% Isn't a Fluke. It's the Job Description.
Stanford's Digital Economy Lab called young workers "canaries in the coal mine" in the August 2025 analysis by Erik Brynjolfsson and co-authors. They used ADP payroll records from millions of US workers and found a 13% relative employment decline for 22-25 year olds in the occupations most exposed to generative AI, while older workers in the same occupations stayed flat or grew. In software development, where AI gains are clearest, employment for 22-25 year olds fell nearly 20% from 2024.
Marketing sits in that AI-exposed bucket for a simple reason: the work juniors were hired to do is the work AI models are shipped to automate. Writing ad copy, drafting landing pages, pulling reports, running first-draft creative briefs, basic keyword research. What's left after a competent AI pass is mostly review, judgment, and strategy. Those are not the tasks anyone hires a 22-year-old for.
This part surprised me when I first saw the Stanford numbers. The decline is not hypothetical forward-looking stuff. It is already sitting in the payroll records, and marketing functions are disproportionately represented in the "high AI exposure" buckets they used.
Why Marketing Takes a Heavier Hit Than Most Functions
Paul Roetzer, founder of Marketing AI Institute, made the bluntest version of the point on his weekly podcast, cited in MarTech's reporting: "the gap between what AI can do and what entry-level workers have typically been asked to do" has effectively closed. His own company, SmarterX, is growing fast and still can't find roles for the entry-level hires he would like to bring on. His take on execution roles was sharper: "If you were just executing (building landing pages, writing copy, etc.), you're cooked. That is not a job one to two years out."
Boston Consulting Group's Julie Bedard put a harder number on it. 90% of marketing manager tasks, not entry-level, manager tasks, face disruption at a skill level from AI. Bedard leads BCG's People Agenda of the Future initiative and sits on their global AI leadership team. If the middle of the ladder is shaky, the bottom rung being missing stops being an HR detail and becomes a structural planning problem.
Anthropic CEO Dario Amodei has predicted AI could wipe out 50% of entry-level white-collar jobs inside five years. From what I've seen in actual agency hiring plans, Amodei's number still sounds high. But you don't need 50%. The Stanford 13% seems to be enough to break the pipeline on its own if it sticks for another two or three years.
The Senior Shortage Shows Up Around 2030
The load-bearing question most marketing leaders are not asking: where does your 2031 senior hire come from?
The old model was something closer to GE's 30-year leadership program. Hire someone at 22, give them reps across three functions over a decade, and by their mid-thirties they turn into a VP who actually understands the mechanics of the business. It is not nostalgic. It is just math. Senior marketing talent in 2031 is made from junior marketing talent hired in 2025 and 2026. Skip the bottom of the funnel now, and the middle of the funnel dries up in five years. Not ten.
Content Marketing Institute flagged the same pipeline risk in their "missing generation" analysis, and staffing leads across multiple agencies are already reporting that 2029-2030 plans cannot be credibly filled at the mid-level. None of the surface indicators are screaming yet, because senior marketers are still plentiful and relatively cheap. That isn't a sign the pipeline is fine. It just means the damage is five years out, so it falls off the quarterly earnings radar.
And to be fair, this isn't entirely new. Consulting and finance have had the same "we don't train juniors anymore" debate for a decade. What seems different now is the speed. The gap that used to take fifteen years to show up is compressing into five because AI is eating the skill-building work, not just the drudgery.
The Agency Retainer Math Makes This Worse
Most agencies I've watched over the last two years are quietly paying 30-40% premiums for senior specialists. NMS covered Brainlabs willing to pay $260,000 for an SEO lead while most client retainers cap at $5,000 a month. The arithmetic only works if you load up the junior layer to offset the senior cost. If AI displaces the juniors and seniors keep getting more expensive, the pyramid stops balancing.
What agencies seem to be doing instead, from what I can see, is quietly shrinking the pyramid. Fewer juniors per senior. More AI tools as the "junior replacement." The short-term P&L looks fine. The 2029 staffing plan looks impossible.
The Three Cheap Fixes Worth Doing This Quarter
I think most teams will overcomplicate the fix. They will run a big AI training program, hire an AI ops lead, redesign the org chart, commission a consulting deck. Personally, I would do three cheaper things first before any of that.
1. Set a junior-to-senior ratio floor. If you run more than three seniors per one junior, you are not training anyone. Keep at least one junior per three seniors on every pod and budget the junior seat as training cost, not production cost. Pull it out of the utilization target.
2. Rewrite the junior job description around oversight, not execution. The unit of junior work is now: run the AI draft, audit the output, tune the prompt, own the signal back to the model. If the JD still reads "write ad copy" or "build landing pages," you are hiring for a role that won't exist in eighteen months and you're wasting the hire's career in the process.
3. Rotate juniors through every AI workflow in the first six months. The old rotation was across functions. The new rotation is across AI-plus-function combinations. PPC with AI-generated copy. SEO with AI-generated briefs. Paid social with AI-assisted scheduling. Pick your stack, document what the human catches that the model misses, and keep the log. That log is the training asset you would have paid six figures for a decade ago.
The benchmark I would set on fix three: a junior hire should catch at least one meaningful AI output error per week by month six. Wrong stat, wrong tone, misread brief, fabricated citation, miscategorized audience. If they are not catching anything, the review loop isn't real and you are just rubber-stamping synthetic content. That is its own risk, and NMS has covered the governance gap on that separately.
The Political Problem Hiding Under the Capacity Problem
The math here isn't complicated. It is just uncomfortable, because the people deciding whether to fund a junior hire today are usually a generation older than the people whose careers depend on it, and a CMO who cuts juniors now is retired or rotated before the senior pipeline actually breaks. The cost lands on the next CMO, not the current one. That seems to be why so little is being done about it yet, even by leaders who openly acknowledge the problem.
I don't think the 2031 winners will be the biggest teams or the ones with the most AI tools. They will probably just be the ones whose CMO had enough interest in 2026 to protect a junior headcount that looked like a luxury on the P&L. Everyone else will be bidding against each other for a senior pool that never got refilled.
Notice Me Senpai Editorial
