The One-Two Punch That’s Kept Entry-Level Postings Lagging Behind

When you’re just starting out in the labor market, it can feel impossible to even get started. You just need to find that first rung on the career ladder—but in today’s labor market, that rung often feels out of reach.

It hasn’t helped that the labor market has been all over the place for the entire 2020s. After the pandemic—when seemingly nobody was hiring—came the whiplash in 2021 and 2022, when the labor market overheated. Then that surge stopped, many companies went through layoffs, and we entered a “no-hire, no-fire” labor market, in which more workers have stayed put in their roles and fewer companies are looking to add headcount. That’s a challenging market for new graduates to enter, and it feels like there are fewer opportunities than there once were. The Conference Board reports that the share of consumers reporting that jobs are ‘hard to get’ rose to 21.9% in September, up from 20.3% in August.

But as it turns out, entry level jobs are not lower than they were in 2019. The reason it feels that way to many people is that postings for other jobs are higher. For every job (at all experience levels) posted in 2019, there are 1.27 jobs available now. But for entry-level jobs specifically, it’s still just that one job. 

While the number of entry-level jobs has stayed the same, their distribution has changed—meaning an easier road for some graduates and a harder one for others. For instance, healthcare has seen a huge increase in the number of entry-level roles available, while IT and Computer Science has seen a sharp decrease. While all of these jobs require a bachelor’s degree, the greatest gains are generally in hands-on, physical work, while desk-based information work is more at risk. 

Overall, though, the point remains that the market for entry-level workers is softer than the overall labor market. Three research papers, all using Lightcast data, have attempted to answer why. They found two causes: one, perhaps predictably, is generative AI. The other is remote work.

In The Papers

In “Labor Demand in the Age of Generative AI: Early Evidence from the U.S. Job Posting Data,” authors Yan Liu, He Wang, and Shu Yu looked at Lightcast postings before and after ChatGPT’s launch, comparing occupations with higher and lower potential for AI to replace workers. By mid-2025, postings in higher-risk occupations were down 12% relative to lower-risk occupations, with especially large declines in jobs requiring no advanced degree (18%) or little experience (20%). Their findings suggest generative AI may be contributing to fewer entry-level openings.

“Generative AI and Posted Labor Demand: Volume, Seniority and Timing” by Fabio Cerina, Simone Nobili, and Marco Rosso approached a similar problem from a different angle: in order to determine which jobs were “highly exposed” to AI—in other words, which jobs AI was most likely to affect—they used a comparison that included actual LLM use. (The first paper judged exposure by AI’s likelihood to replace tasks or jobs, not its actual usage.)

Studying 368 million Lightcast job postings, they found that postings in the most exposed occupations fell about 8.6% after ChatGPT’s release, but the authors find the decline in junior postings began during the COVID recovery—suggesting AI reduced hiring overall, but the shocks that hit the entry-level market came earlier, in 2021. During the post-COVID hiring boom, remote work surged, and when the authors identified which jobs were best-suited to remote work, they found that this accounts for the gap between entry-level and senior postings. 

That aligned with what Peter John Lambert and Yannick Schindler found in “The Broken Ladder: AI, Remote Work, and Early-Career Hiring.” When analyzing new hires and Lightcast job postings across the US, UK, Canada, and Australia from 2017 to 2025, they found that both GenAI and work-from-home exposure are associated with fewer junior hires when studied separately. When accounting for both links together, though, the link with remote work remains, while the GenAI effect shrinks and often is no longer statistically clear—suggesting remote work may explain more of the decline in early-career hiring.

A Harder Ladder To Climb

None of these papers provide hard causal evidence, but from their analysis of Lightcast job postings, it seems like remote postings were the catalyst that pushed entry-level postings down relative to others, and then the arrival of generative AI kept them from recovering. Shot, then chaser. 

What does that mean on a practical level? With fewer entry-level jobs available, employers have more leverage in what they ask of these new hires. 

In an article for Fortune, the former president of the Federal Reserve of Philadelphia made the case that companies hired entry-level workers with the implicit understanding that they were training their future leaders, without a real expectation of productivity in their early years. Their routine, low-level work (like summarizing documents and preparing briefs) was essentially on-the-job training. But now, since many of those tasks can be automated for less than the cost of an entry-level salary, companies are pulling back and being more selective and demanding of these workers. (The author cites yet another paper that uses Lightcast data: “Looking for the Ladder: Is AI Impacting Entry-Level Jobs?” by Zanna Iscenko and Fabien Curto Millet.)

Our own Lightcast analysis shows employers have begun to request higher-level leadership and management skills for these jobs. Instead of workers being trained for operations and planning over the years, they’re expected to come in already possessing those skills. 

If a career is a ladder, then the first rung is getting both narrower and taller: fewer entry-level jobs are available in many career areas, and the expectations for entry-level workers are getting higher.

Thanks for reading On The Job. Be sure to catch up on our past issues, including "Oil, Energy Independence, and the Workforce Gap” "Tracking The Agentic Explosion" and "Degree Requirements Are Falling, Except for AI Jobs." You can also subscribe here. We’ll see you next time.