“It was very devastating,” he told Fortune. “Honestly, I thought that having a 3.98 GPA, getting recognition letters, and having an interesting experience in the past, perhaps I could get a full-time job offer easily. But that was not true.”

Kang, who lives in Portland, Ore., eventually landed a job at Adidas, where he had interned the previous summer, after more than 10 months of endless job applications. His experience is actually better than that of many of his fellow grads; one of his classmates, he said, has been on the job hunt for two years.

For many new graduates, the first rung of the corporate ladder is getting harder to reach. Entry-level roles, typically defined as positions requiring no more than one year of prior full-time experience and providing on-the-job training, are becoming increasingly rare in many white-collar industries.

Job postings are down, internships are converting to fewer permanent roles, and some employers now expect entry-level hires to arrive with skills once taught in-house.

Artificial intelligence is accelerating this trend by automating junior-level tasks and giving companies an incentive to delay or reduce early-career hiring. Experts warn that while this may cut costs in the short term, it could weaken the leadership pipeline in the years ahead. In the tech sector, hiring for new graduates in the 15 largest companies fell by over 50% since 2019, according to a report from VC firm SignalFire. Before the pandemic, new graduates made up 15% of Big Tech hires, now, that number has dropped to just 7%.

“I feel like it’s getting worse over time,” Kang said. “AI is taking over, which is creating limited jobs or just pushing companies to look for very high-level candidates. I feel like it’s very unfair.”

A weak training pipeline

For the past few decades, climbing the corporate ladder has been a relatively straightforward process. Graduates started with an entry-level role where companies invested in training and development before steadily promoting from within.

Truly entry-level roles have been disappearing for some time as companies increasingly expect new hires to arrive with skills and experience that once would have been taught on the job. But as the AI efficiency drive eats away at even more entry-level roles, the hiring market for new grads is getting difficult.

Companies are being more selective about who and where they hire while they attempt to integrate AI, and entry-level roles are feeling the worst of this impact because the technology is particularly effective at automating tasks handled by junior workers, such as data cleaning, summarization, and basic QA.

According to data from Handshake, a Gen Z-focused career platform, entry-level job postings for traditional corporate roles decreased by approximately 15% last year.

Internship conversion rates are also slipping. In 2023–24, only 62% of interns received full-time offers, pushing the overall conversion rate below 51%, the lowest in more than five years, according to the National Association of Colleges and Employers. Hybrid interns converted at even lower rates than those who were in person.

One of the ways to succeed in the age of AI is to leverage expertise to work productively with more generalized AI tools. But if traditional career trajectories run dry, what happens when companies run out of experts?

“I’m sure there’s going to be big skills gaps,” said Stella Pachidi, a senior lecturer in technology and work at King’s Business School. “I think that the traditional ways in which we have seen people developing expertise could easily vanish.”

Studies are increasingly pointing to AI as one of the drivers behind the shrinking job market, particularly for entry-level roles. In the U.S., in the first seven months of 2025 alone, AI was cited as a reason for just over 10,000 job cuts, according to new data from outplacement firm Challenger, Gray & Christmas. The firm ranks AI among the top five stated causes of workforce reductions this year.

These disruptions could cause problems down the line.

“If a lot of firms are cutting, cutting, cutting at the entry level, there’s a fear that they might actually miss out on the talent that’s going to create their pipeline going forward, that’s going to become the managers, executives, etc.,” Tristan L. Botelho, associate professor of organizational behavior at Yale School of Management, told Fortune.

“Everyone is just focused on the current efficiencies and not necessarily thinking further about the future,” Pachidi added. “How will their organizations be doing? What kind of value will they be creating in the future, and will they have experts?”

Concern about tomorrow’s talent

A looming skills gap is something a lot of executives are worrying about, according to Nick South, managing director and senior partner at Boston Consulting Group.

Although he sees the disruption of entry-level jobs as a “short-term” issue and believes that long-term AI will actually create new jobs, this brief disruption could be a painful one.

“At the point in time for an individual, this is incredibly disruptive, and as a society, we need to help people with reskilling,” he said.

There’s also the question of how to prep young people for an AI world. Some argue that the rise of AI might rewrite the traditional path from education to entry-level jobs entirely.

“The middle ground of knowledge workers is likely to become less important,” said Rob Levin, McKinsey senior partner and a leader of QuantumBlack, McKinsey’s AI arm. “And I worry about how we are going to incent folks to deeply specialize and get companies to train folks in those deep specialties that they need. Will there be new vocational schools or things?”

At universities, professors and students have both realized that the majority of work performed in some university courses can be assisted, if not near totally automated, by AI.

Students were some of the first to realize ChatGPT’s ability to write essays and to summarize long texts. But while students may find their academic load is significantly lightened by AI tools, professors told Fortune they were worried about the prospect of a generation lacking critical skills and traditional education.

A study from MIT suggested that LLM (large language model) use can reduce neural engagement and harm learning in students, especially younger users (researchers caution that the findings are early and not yet peer-reviewed). The study also found that ChatGPT users specifically had the lowest brain engagement and “consistently underperformed at neural, linguistic, and behavioral levels.”

Graduates taking action

Eva Selenko, a professor of work psychology at Loughborough Business School, thinks that educational systems and the job market are more likely to adapt to a generation of AI-boosted talent than to cope.

“I think we need to educate people to use AI tools to the best of their expertise,” she said. “I do absolutely feel for those graduates. On the other hand, you know, they are super, super, highly educated people. They have drive, they have creativity.”

Young job-seekers are already taking some of this upon themselves.

While job hunting, for example, Kang founded a startup as a way to gain experience, since employers were requiring years of experience even for entry-level positions. He formed a group with other computer science graduates in similar situations to do tech consulting for clients at low cost, to build their résumés.

“I’m not just sitting here applying for jobs and just biting my nails,” he said. “I’m here to continually do other activities along the way.”