Degree Requirements are Dropping—But They’re Still Higher for AI Jobs

It’s no secret that at Lightcast, we like skill-powered hiring. This is the process where employers hire based on the specific skills they need in the job, rather than broad credentials such as a four-year degree when they aren't necessary for the work. Not only does this clarify job requirements on both sides—so that individuals can be specific about the skills they can offer, and the company can be specific about what they need—it also helps expand the potential talent pipeline, so that more workers can access well-paying jobs, whether or not they have a four-year college degree. 

But if there’s one thing we’ve learned over the past few years, it’s that AI complicates things. And in this case, it might be for the better: generative AI arrived too quickly for traditional credentials to keep pace, which might be in argument in favor of skills. If employers suddenly needed capabilities that experienced workers couldn't have studied in college, perhaps demonstrable skills would matter more and degrees would matter less.

So has that happened? Let’s look at what the data says. 

Degree Requirements are Falling Overall

Using 2019 as our starting point (baseline number before the pandemic-era labor market disruptions and before mainstream AI launched), we can look at which jobs traditionally require four-year degrees and track how much those jobs have changed their degree requirements. We can use this as a proxy to track skills-powered hiring: a lower share of jobs requesting degrees would imply more employers using skills in their hiring.

In fact, that’s exactly what we see. Degree requirements are down roughly 14% since 2019 in US jobs that have typically required college degrees. (To be clear, this doesn’t necessarily prove that employers are looking for skills specifically to fill these jobs, only that they’re adapting their hiring strategy in some way that does not necessitate a four-year degree). 

But while a lower share of jobs overall require the same education, that exact figure varies dramatically based on career area. Transportation, for instance,  has seen degree requirements decline 24%, but on the other hand, Law, Compliance & Public Safety has seen an increase of roughly 19%. 

A Job Requiring AI Skills Is More Likely To Require A Degree

So skills-powered hiring is increasing at the same time that AI has swept through the labor market. Is there a connection? Can we attribute the drop in degree requirements to broader AI use?

As it turns out, we can’t, because AI is correlated with higher degree requirements. Among traditionally college-educated jobs, postings that require AI are actually more likely to require a four-year degree than those that do not.

Across many different career areas, AI postings have higher degree requirements, and have consistently for years. (The only exception is Community and Social Services, where in 2025, 54.8% of the total jobs required degrees, compared to 54.7% of AI jobs—effectively tied.) In some fields, such as that one, the gap is shrinking, while in others, like Healthcare, it’s widening, but in almost every instance, AI jobs have consistently required degrees more than non-AI jobs. 

This finding requires some nuance. Lightcast defines an AI job as one that requires at least one of the 300+ AI skills Lightcast has identified (the same methodology used to power our analysis for the Stanford AI Index every year). That means that many of what we call AI jobs—in fact, over 50%—are actually not in the tech industry. When we do look at jobs in IT and Computer Science, the relationship between AI and degree requirements gets more complicated. 

AI Jobs in Tech Defy Expectations

Just like the other career areas studied above, AI jobs in tech are more likely to require a college degree. But beneath that top-level finding are some surprising wrinkles. 

This is an area where skills-powered hiring has often been celebrated, because workers need specific skills like knowing a coding language, but in the first chart above, it had the second-highest increase in degree requirements since 2019—even as overall degree requirements declined. 

We can separate IT & Computer Science jobs into three tiers: AI engineers would have the most expertise, followed by non-engineer AI jobs, and finally non-AI jobs. The chart below tracks degree requirements for all three segments, as well as the overall average for all IT & CS jobs. 

For both AI engineers and AI jobs overall, degree requirements are down since 2019. It hasn’t been a straight line down for either category, but there’s been a clear decline since 2024. The overall and non-AI trendline has been more consistent in its increase since 2019. 

Out of all these data points, the decline in degree requirements for AI jobs makes the most intuitive sense. Although we know that AI workers, especially AI engineers, are highly educated, the degree itself might be less of a prerequisite for their jobs than their actual expertise. But if that were true, the same should be true of non-AI workers in the tech space as well, and it isn’t. 

So why have employers retained degree requirements for emerging skills that often must be learned after graduation?

There are several possible explanations. In many fields, AI may be an additional skill layered on top of deep domain expertise. A healthcare, finance, or engineering job that requires AI skills may still require the education associated with the underlying profession. The mix of employers and jobs could matter as well: companies recruiting for AI skills may be hiring for more specialized or senior positions, or may differ in other ways from employers recruiting for non-AI roles.

Or the answer might be simpler, and more frustrating, than that. Employers might be adding new AI skills to existing jobs without reconsidering their traditional requirements, even in the face of changing circumstances. In that case, the persistence of degree requirements would reflect the way hiring practices evolve: skills can change faster than the systems employers use to evaluate them.

This analysis can’t tell us which of those explanations is driving the pattern, but it does tell us that AI is not automatically removing credential barriers. Employers that want to practice skills-powered hiring still have to make deliberate choices about which credentials are actually necessary—and those choices may look very different from one occupation to another.

AI may be changing the skills employers need. But that won’t matter if employers don’t change their approach to talent in response.