The Perils of Using AI to Replace Entry-Level Jobs

As you’ve probably heard, if not also experienced, AI is reshaping how we work. If you happen to be a senior professional with sufficient reputational capital, status, and a deep social network, you may be relatively safe from being displaced by AI—at least for now.

However, there is mounting evidence that the same is not true of entry-level jobs. A Stanford study found that U.S. employment for early-career employees in the most AI-exposed fields, such as software development and customer service, has fallen substantially in recent years. Research at the World Economic Forum suggests that 50% to 60% of typical junior tasks (report drafting, research synthesis, coding fixes, scheduling, data cleaning) can already be executed by AI.

Why Organizations Should Redesign Entry-Level Jobs

We believe slashing entry-level jobs simply to cut costs is dangerously short-sighted—both for companies and society. There is a strong case to be made that organizations must resist the temptation to eliminate them en masse; instead, they should redesign them. There are at least four compelling reasons:

1. To build future mid-level professionals and leaders

Every capable manager and professional starts somewhere. The best leaders and the most impressive experts acquire the skills and perspectives they need to lead and solve important problems by learning the trade from the ground up. In any profession (or activity), over time people shift from being consciously incompetent to competent, and ultimately to becoming unconsciously competent in a set of valued skills. At this point, they are better able to see the larger picture that supports making high-stakes decisions effectively. Stripping out entry-level jobs severs this pipeline.

Advertisement

Imagine recruiting managers who have never worked at the front lines, never handled customer complaints, never written up notes from consequential meetings, never grappled with the minutiae of operational work. Leadership would become abstract, detached, and dangerously naive. Recall, if you will, the eureka moments in your own career, where the recurrence of similar patterns allowed you to glimpse an important cause-effect relationship—perhaps cementing an insight that made you more effective in your work or profession.

Simply put, most of today’s entry-level jobs accomplish two essential functions: They get tasks done, and they develop the people doing the tasks into more capable members of the organization—and society.

2. To fuel innovation from the ground up

Innovation often bubbles up from those closest to the work. Junior employees, unencumbered by legacy thinking, are uniquely well positioned to spot inefficiencies and suggest creative fixes. In technology, this is called “dogfooding”—as in, eating your own dog food. Microsoft famously tested early versions of Word and Excel internally, using staff feedback to shape the products before public release.

Entry-level workers generate similar value by stress-testing processes and discovering what is broken. Unlike AI, which delivers consistent outputs, humans introduce variability, which is sometimes messy but often the source of new ideas, improvement suggestions, and occasional breakthroughs. If your approach to innovation is to outsource ideation to the same machine or tool everybody can use to produce a very similar or even identical outcome, don’t expect to produce a competitive advantage.

3. To enrich the organization’s culture

Today’s workforce is more intergenerational than ever, with up to five generations working side by side. This diversity of age and perspective enriches workplace culture and sparks creativity. Removing young people drains organizations of fresh energy and narrows the range of viewpoints. It risks producing sterile, homogeneous cultures where established employees talk mostly to each other and the cycle of renewal breaks down.

4. To protect society

Work is more than income; it provides purpose, structure, and belonging. Without entry-level roles, millions of young people risk drifting into idleness. We know from history that when large groups of able-bodied young adults lack meaningful occupation, societies pay the price in the form of alienation, unrest, and even crime. Protecting entry-level jobs is not just a corporate responsibility but a civic one.

How to Redesign Entry-Level Jobs

To protect entry-level jobs, they must be reimagined so that they still deliver value in an AI-powered workplace. This involves four major steps:

1. Redesign tasks.

Junior roles must no longer be defined by the repetitive, automatable tasks that AI can do better and faster. Instead, they should be designed to expose people to the why behind the work.

Take accounting. AI can reconcile transactions or draft financial statements, and most accounting firms have been deploying various forms of AI to redirect junior staff into higher-value activities like anomaly detection, fraud investigation, and client advisory work. This allows junior staff to still learn the mechanics, but their real work is to interpret what the machine produces.

Likewise in the world of staffing and high-volume recruitment where one of us (Tomas) has worked, AI has been used for years as a tool that enables human recruiters to prioritize high-potential job seekers, sifting through millions of CVs, application letters, and prerecorded interviews. However, people, even junior recruiters, still play a key role in the “last mile delivery” of the value chain by using the time they save not doing tasks outsourced to AI to conduct more valuable human-to-human exchanges with shortlisted candidates and clients.

These kinds of shifts align with McKinsey’s estimate that, while 60% of occupations could see at least a third of their tasks automated, very few can be fully automated. The real opportunity lies in rethinking jobs so humans spend more time where judgment, collaboration, and creativity are needed. In this way, human aptitudes are amplified by AI.

2. Focus on augmenting skills.

AI is only useful when paired with critical thinking. Productivity gains are meaningless if they come at the expense of professional judgment. A recent study published in Science found that generative AI can boost output by as much as 40% in text-based tasks, but novices who accept the machine’s suggestions uncritically perform worse than those who reason through problems themselves.

Consider banking, where most analysts use generative AI to draft presentations and reports. To complement this, training now includes “red teaming” exercises where juniors are asked to test assumptions, identify weaknesses, and explain why the gen AI might be wrong. The goal is not speed alone but also judgment. The early-career analysts are asked to interrogate the AI’s output the way a skeptic or competitor would—to probe for incorrect assumptions, missing data, or logical flaws—and then defend their critique to senior colleagues. This turns the AI into a kind of intellectual sparring partner: fast and capable but fallible.

3. Redesign work.

The default use of AI is substitution: Let the machine do the work and cut headcount. A smarter approach is to redesign workflows so AI handles rote execution while humans focus on framing the problems, asking better questions, and building relationships.

In consulting, AI can synthesize market reports. But firms still embed junior consultants in workshops and interviews, where they develop interpersonal skills and a sense of context that no algorithm can provide. In software development, gen AI platforms are widely used to write boilerplate code, but junior engineers are steered toward debugging, system design, and pair programming (a collaborative technique for software development)—areas where collaboration and problem-solving matter most.

So far, most of the research on hybrid human-AI workflows suggests that the highest performance comes not from “AI first, humans second” but from a carefully structured division of labor where machines accelerate routine work and people focus on uncertainty, novelty, and persuasion. This is consistent with Ravin Jesuthasan and John Boudreau’s research on “work without jobs,” which shows how organizations can move beyond “job titles” and “job holders” to adopt more fluid, skills-centric, and work-driven operating models.

4. Develop people.

Perhaps the most important principle is that entry-level work should be designed not just to get things done but also to develop people. Early exposure to pressure, ambiguity, and even failure is how professionals acquire resilience and judgment.

Consider medicine. Residents still endure long, exhausting shifts. AI may eventually automate charting or scheduling, but hospitals keep juniors on the front lines because these experiences build clinical intuition and empathy under stress. In journalism, AI can draft articles, yet young reporters still get sent to cover tedious community meetings or chase cold leads, because persistence and interviewing skills only come through practice.

Equally important is rediscovering the value of resilience and grit. If machines remove every obstacle, work becomes too easy, devoid of the challenge that makes learning meaningful. As one of us (Amy) has argued in her research on failure, progress comes from intelligent failures: the false starts, stumbles, and disappointments that occur when tackling difficult, uncertain tasks. Entry-level jobs, by providing safe spaces to try, fail, and try again—where the stakes are lower than at the top—are vital for building adaptive and confident professionals.

Consider the analogy of education: If a student outsources every essay to generative AI, he bypasses the intellectual struggle that produces deep learning. It is like microwaving ideas: fast, convenient, and unsatisfying. The effort, even the pain, of thinking for yourself is what grows a student’s capacity. The same applies in work. If we remove the stretch and discomfort of early jobs, we rob future leaders of formative growth that allows them to tolerate the stretch and discomfort of leadership roles.

Beyond the Organization: A Societal Shift

Almost 100 years ago, John Maynard Keynes predicted that by the year 2030, technological progress would reduce the workweek to 15 hours. Although time may prove him right, people still argue today about whether (gasp) even a four-day work week is too short. The evidence for the benefits (in productivity, well-being, and retention) of the four-day week is robust, but organizational norms (and mindsets) lag behind.

AI presents an opportunity to reset. Instead of extracting maximal “rent” from every worker, as if they were hourly machines, organizations could redefine value more holistically: quality of output, contribution to culture, and capacity for innovation. That requires not just fewer hours but smarter, better work, and entry-level roles are the foundation for where that shift must start.

The instinct to automate away entry-level jobs is understandable but short-sighted. These roles are not inefficiencies to be eliminated; they are investments in the future of leadership, innovation, culture, and society itself. As AI transforms work, our task is not to minimize human involvement but to maximize its value. And that starts with protecting, redesigning, and re-dignifying entry-level jobs.

For what it’s worth, part of the challenge—and part of the reason professors haven’t significantly adapted their courses to the seismic shifts created by AI—is that no one really knows how this will all play out. If the ultimate purpose of higher education is to prepare people to be effective contributors in the world, then it’s profoundly difficult to do so without a crystal ball showing us what tomorrow’s world will require. Today’s content may soon be obsolete, but the habits of effort, self-discipline, and systems thinking that people will need to thrive in the future transcend the specific content being taught today.

The same is true for companies. Organizations are operating without certainty about what roles, skills, or even business models will define success in five or 10 years. None of us knows enough about the future we are preparing for to get it entirely right. That is why the old saying resonates more strongly than ever: The best way to predict the future is to create it.

And that is precisely the imperative in front of us. Protecting entry-level jobs is not about defending tradition but about actively shaping a future where work remains a site of growth, resilience, and shared human achievement. If we seize this moment with imagination and courage, the AI age can be not the end of opportunity but the beginning of a smarter, fairer, and more fulfilling world of work.

Copyright 2026 Harvard Business School Publishing Corporation. Distributed by The New York Times Syndicate.

Explore AAPL Membership benefits.