Give Youth an AI Chance

The data is in. AI is quietly closing the door on the first rung of the career ladder — and most companies are pretending not to notice.

There is a number sitting in the payroll records of millions of American workers that every chief executive, every hiring manager, and every parent of a recent graduate should read slowly.

Among workers aged 22 to 25 in fields most exposed to AI — software development, customer service, accounting, sales — employment has fallen 13% since late 2022. That is not a forecast. That is not a model. That is what happened, measured from the actual payroll records of millions of workers at thousands of companies, provided by Automatic Data Processing (ADP), the largest payroll processor in the United States, and analyzed by economists Erik Brynjolfsson, Bharat Chandar, and Ruyu Chen at the Stanford Digital Economy Lab.

The paper is titled "Canaries in the Coal Mine?" Published in August 2025, it is one of the most carefully constructed labor market studies of the AI era. And its findings are not comfortable.

In software development specifically, entry-level employment for workers aged 22 to 25 was down 20% from its late 2022 peak by July 2025. In customer service, nearly 11%. Older workers in the same fields — people aged 30 and over — saw employment grow between 6% and 9% over the same period.

Brynjolfsson explained the asymmetry plainly: "These large language models are trained on books, articles, and written material — that is the kind of book learning that a lot of people get at universities before they enter the job market. So there is a lot of overlap between these large language models and the knowledge young people have."

AI is not replacing careers. It is replacing the entry point to one.

The numbers keep getting worse

In April 2026, Goldman Sachs published research finding that AI is erasing approximately 16,000 net U.S. jobs every month — 25,000 positions eliminated by AI substitution each month, partially offset by 9,000 jobs added through AI augmentation. Goldman tracked 21,900 layoffs explicitly attributed to AI in April 2026 alone, the highest single-month figure since the bank began counting in 2023. Total AI-attributed layoffs now stand at 136,000 over three years.

Tech layoffs through May 2026 have already passed 115,000 — approaching the 124,000 logged in all of 2025, with Meta, Amazon, Block, and Snap among those citing AI as a driver of cuts.

The entry-level job market has collapsed in ways that go beyond tech. According to a Guardian survey of 850 business leaders across seven countries, 41% say AI is allowing them to cut staff, and 25% say entry-level tasks could soon be performed primarily by AI. Job postings on Handshake, an early-career talent platform, fell more than 16% between August 2024 and August 2025. The average number of applications per role jumped 26% over the same period.

58% of Gen Z students who graduated in 2024 and 2025 were still looking for their first job, compared to just 25% of millennial and Gen X graduates in previous years. Youth unemployment in the United States reached 10.8% in 2025, according to the U.S. Bureau of Labor Statistics — more than double the overall rate.

And this summer may be the worst on record for young people entering the workforce. U.S. employers are expected to hire teenagers for approximately 790,000 jobs this summer — lower than any year on record. On Indeed, job postings for summer interns and camp counselors are both down more than 25% year over year. Roughly 20% of companies say they will hire fewer interns or forgo them completely this year.

This is not a blip. It is a structural shift — and it is accelerating.

Why this should alarm every business leader, not just young people

When companies stop hiring at the bottom, they do not just hurt young people. They hollow out their own pipeline.

The junior analyst of today is the senior manager of 2032. The entry-level developer of today is the engineering lead of 2035. Cutting early-career hiring may look efficient in a quarterly earnings call. It looks catastrophic five years later when you have no one to promote, no institutional knowledge coming up from below, and a talent acquisition bill that is two to three times higher than it needed to be.

International Business Machines (IBM) chief human resources officer Nickle LaMoreaux made this case directly in February 2026 when she announced that IBM would triple its entry-level hiring in the United States — including specifically for software developers and other roles AI is said to be replacing. Her reasoning was not charitable. It was strategic.

"The companies three to five years from now that will be the most successful are those that doubled down on entry-level hiring in this environment. Not those that held the status quo or reduced it."

IBM has not kept the old job descriptions. It has rewritten them entirely. Junior developers who once spent 34 hours a week on routine coding now spend that time working with customers, collaborating with marketing, and building new products. AI does the repetitive work. Humans do the work that requires a human.

That is not a soft human resources position. It is a business model.

Two more companies getting this right

Box, a cloud software company serving more than 65% of the Fortune 500, told The New York Times it has never conducted broad-scale layoffs and has no intention of starting. Since integrating AI, it has created entirely new job categories that did not exist two years ago: AI model evaluators who benchmark which tools perform best for specific tasks, automation engineers who help internal teams adopt AI, and forward-deployed engineers who work directly with customers who lack the technical capacity to implement AI on their own.

Chief Executive Aaron Levie's argument is simple: when AI makes your teams more productive, you can pursue markets and projects that were previously too expensive to justify. New markets require new people. The companies treating AI as a cost-cutting tool are solving the wrong problem.

Schneider Electric, a French energy technology company with 160,000 employees worldwide, offers perhaps the clearest proof of concept. Chief AI Officer Philippe Rambach built an augmentation-first strategy across the company's global workforce. At a manufacturing facility in Normandy, AI systems now control chemical washing cycles with precision that reduced production waste by 73%. In call centers, AI searches through documentation to surface answers for agents who then apply judgment before responding. In the final quarter of 2025, AI provided the correct response for 75% of straightforward customer questions — freeing agents to handle the complex cases that actually require a person.

Nobody lost their job to make any of it happen.

The Stanford study draws the distinction that ties all of this together. Not all AI deployment leads to job losses. The negative effects are concentrated in fields where AI is replacing human tasks. Fields where AI is amplifying what humans can do are not showing the same entry-level declines. The difference is not the technology. It is the decision companies make about how to use it.

The question that matters

Gen Z excitement about AI collapsed from 36% to 22% in a single year, while Gen Z anger about AI rose from 22% to 31% (Gallup and Walton Family Foundation, 2026). Only 10% of Americans say they are more excited than concerned about AI, according to Pew Research in March 2026.

That is not a communications problem. That is a trust problem. And companies created it.

A workforce with no pipeline ages out. A culture that stops bringing in young people stops learning how they think, how they work, and what they expect from an employer. And in a moment when AI-fluent young people may be the fastest path to organizational transformation — as Levie, LaMoreaux, and others have argued — cutting them out of the workforce is not just harmful to them.

It may be the most expensive mistake a company makes this decade.

The question for every leader reading this is simple: Will you show that you are on the right side of history?