The end of 2025 painted a dour picture of the job market, especially entry-level roles.
Entry-level jobs are disappearing, says one survey. Companies plan to replace those positions with artificial intelligence tool use, says another. The tech may even be killing younger workers’ job prospects at a disproportionate rate.
But amid a market recently described as “frigid,” is AI actually to blame?
The story is more nuanced than that, experts told HR Dive, but the ramifications of AI adoption are numerous. And employers that use the technology for tasks typically apportioned to entry-level employees may be unprepared for the changes such investment may require.
What the narrative hides
Some high-profile layoffs and job cuts at the end of 2025 — including 14,000 roles eliminated at Amazon — were pinned on AI adoption, but that may be a convenient narrative for employers, Jason Walker, co-founder, CEO and CPO of Thrive Consulting, told HR Dive in an email.
“Layoffs are pretty much always on the table, frankly,” Walker said. “But I think what makes them easier to do is the guise of ‘we need to get more efficient’ and AI allows for that narrative.”
Companies are hiring less, he continued, particularly right now as they assess the effect of AI and whether it will actually make them more efficient.
Entry-level workers seem to be at risk, but they aren’t necessarily the only victims of cuts, and they’re not necessarily being laid off faster than anyone else, Emily Rose McRae, senior director analyst with Gartner, told HR Dive. But organizations are hiring less, she said, and universities report difficulty helping graduates land jobs.
Early career workers aren’t likely to be completely shut out from the market, either, Wendy Mazoway, global vice president of recruitment process outsourcing at Korn Ferry, told HR Dive. One of the company’s clients, a large global aerospace tech manufacturer, brought AI into the fold while maintaining a “very strong commitment to hiring early career folks,” Mazoway said. That program for early hires has actually grown, she added, spotlighting how work is changing but still creating opportunity.
So is AI adoption really driving massive headcount shifts? “The thing is, no one can really tell you how much of that is due to excellent returns in AI,” McRae said. “It probably isn’t at all.”
Employers are still in speculation mode when it comes to the technology’s benefits. Many may be pressured to make cuts as a form of “debt financing,” hoping AI investment gains will eventually justify the layoffs, McRae added.
Uncertainty about AI’s actual impact on operations, however, remains a sticking point. While some companies, such as Salesforce and Amazon, have openly explained how AI is changing their businesses, Walker said, those tangible examples are few and far between.
“Nobody can really tell you tangibly [t]hat AI is going to improve their business and make it more efficient,” Walker said.
AI’s benefits have been notoriously difficult to quantify, Walker said. And as employers experiment, they may struggle to see the powerful productivity gains often promised by tech companies.
The key, according to McRae, is clarity: “Don’t make investments in your talent strategy assuming some sort of tech capability you don’t have access to,” she said.
If entry-level jobs are at risk, what does that mean for pipelines?
Even if AI’s impact on job cuts is so far overblown, the reality is that many entry-level jobs are, in fact, at risk, Walker said. And that effect can be quantified.
Entry-level researchers, marketing communications professionals, administrative assistants and even early software engineers have seen large aspects of their work be taken over by automation, he noted.
Employers also may redesign workflows to better use AI, which could lead to entry-level jobs being cut, McRae said. But that creates another problem: How does an employer build a sustainable, trustworthy pipeline of workers with the experience needed for higher-level jobs?
“That’s a two-fold problem,” McRae said. Not only are workers not obtaining the skills, in that case, but there’s also no easy proxy for competency — such as years of experience in a certain role — that employers can look to when considering promotions or hiring.
And while employers can train those workers up, as Mazoway noted, those painstakingly trained workers could easily be poached by competitors instead — and they often are, McRae said. Too much uncertainty will drive people to leave fields entirely, making the talent crunch even more severe.
A question of change management and decision-making
Because AI is still so new, few guides exist on how to prepare the workforce for its impact, McRae said. To see returns on those investments, employers may have to lead costly redesigns on workflows and job designs — and many leaders, after seeing what effort it would take, may not see the change as worth it, she noted.
Employees, too, may be wary of employers’ attempts at change management. A December report from Eagle Hill Consulting showed that only 1 in 4 employees surveyed said their organizations effectively manage major change.
But not all change spells doom for jobs or opportunity. Many companies are investing in training centers again after they fell out of favor during the pandemic, Mazoway said. AI, in general, will require significant investments in infrastructure that will create jobs, she added.
AI is just the latest technology changing how work is done, and knowledge-based jobs have long faced this trend, she noted. “If I think about my job as a recruiter 30 years ago, I was flipping through a phone book and calling people,” Mazoway said. Instead, the person with that job today is much more focused on relationship building thanks to developments in applicant tracking systems and other recruiting technology.
But generally, change management will remain the elephant in the room as AI conversations move forward, Walker said.
“I think educating and training are not the issues, it is change management and understanding how you use AI to be more efficient,” he said. “That is what people need to learn.”
