Laura Ullrich has sympathy for college graduates looking for work. The director of economic research for the job site Indeed knows that struggle intimately. Her son, a data scientist, is graduating with a master’s degree this year. “Because of the job I do, I get asked by lots of his friends’ parents and friends for help,” she says. “But it’s brutal out there right now.”
The labor market worsened for recent college graduates, those ages 22 to 27, at the end of last year. The unemployment rate climbed to about 5.7% in the fourth quarter of 2025, an uptick from prior months and above the rates of 4.2% for all workers and the 3.1% for college grads of all ages, according to the New York Fed.
Graduates like Ullrich’s son, who are eyeing the tech field, are facing an added hurdle: a phenomenon Ullrich calls “experience creep,” in which employers are seeking higher levels of experience at the expense of opportunities for early-career professionals. The share of postings open to those with two to four years of experience dropped from 46% in mid-2022 to 40% in mid-2025, while the share seeking at least five years of experience jumped from 37% to 42%, according to Indeed data.
The trend, in part, boils down to supply and demand. “The reality is that it’s more of an employer’s labor market, and so they have the freedom and ability to ask for more years’ experience,” Ullrich says. “If you can hire somebody with several years’ experience, why hire an entry-level person?”
The preference for more experienced candidates also aligns with the rise of AI that is capable of doing lower-level work—the kind of grunt work that can often be a way in the door for early-career workers. A November report from Stanford economists found “substantial declines in employment for early-career workers (ages 22-25) in occupations most exposed to AI, such as software developers and customer service representatives” while “overall employment continues to grow robustly.” Together, the results support the idea “that generative AI has begun to affect entry-level employment.”
Ullrich isn’t totally convinced that AI, as a tool, is directly to blame. There are still few smoking guns pointing to employers actually replacing human workers with AI agents. “What is hard to know is how much of this is actually about AI technology disrupting employment versus AI investment disrupting employment,” Ullrich says. There’s growing evidence that it’s the latter, with companies prioritizing capital expenditures over labor in the great, costly AI build-out. Just this week, Oracle laid off scores of workers as it plows billions into building data centers for AI development.
“[Companies] may just be spending less on labor because of that capital-labor tradeoff, just like they would if, all of a sudden, they decided to buy a bunch of new equipment. That’s kind of what always happens when we get through periods of technological disruption,” Ullrich says. “But this is also different, because AI can do some of the work that entry-level folks are doing, so it’s really hard to disentangle those two.”
The experience creep phenomenon is particularly acute in tech, which is the softest of sectors in terms of hiring, giving employers a firm upper hand. U.S. job postings on Indeed for software developers of all levels, for instance, are currently down 29% from Indeed’s pre-pandemic benchmark. Data and analytics jobs are down 38%.
In the short term, experience creep is great news for tech companies that can staff up with more seasoned workers for the bargain prices of entry-level rookies. But it’s a trend that may catch up with employers in the long term. The growth in tech jobs is happening at more senior, higher-paying levels, Ullrich says. “How do you get the number of senior people you need if you’re not training junior people?”
It’s a question all firms will face if AI really does eliminate large swaths of entry-level work, as some CEOs like Anthropic’s Dario Amodei have predicted.
For now, Ullrich is advising young graduates—her son included—to lean into AI and “prove to companies that you plus AI is better than AI without you.”
