AI Skills Are in High Demand, But AI Education Is Not Keeping Up

(Deemerwha studio/Shutterstock)

Millions of workers in the U.S. say they want to reskill in AI. Nearly 47 million white-collar professionals plan to start learning within the next six months, driven by the sense that AI will reshape their jobs, whether they’re ready or not. The urgency is real on the employer side, too. Roles for AI and machine learning engineers are projected to grow by 40% each year. That is nearly 100 times faster than the average for all occupations.

Yet for most people, the path to acquiring those skills remains unclear. Very few learners are going through traditional degree programs, and even fewer are finding formal support for AI skills. Most are navigating it alone, piecing together courses, tutorials, or certificates wherever they can.

This is the backdrop captured in a new report from Validated Insights, which lays out in clear terms what many educators and employers already feel: AI is moving faster than the institutions meant to support it. The report traces the full arc of the crisis, from enrollment numbers to hiring trends.

A key theme that emerges from the report is that the problem isn’t with interest or lack of motivation; it is the infrastructure for learning AI that appears to be too limited in scope and growing too slowly to support the meteoric rise of AI. 

(Inkoly/Shutterstock)

There’s already a big gap between how many AI workers are needed and how many are available, and it’s only getting worse. The report says the U.S. was short more than 340,000 AI and machine learning workers in 2023. That number could grow to nearly 700,000 by 2027 if nothing changes.

Faced with limited options in traditional higher education, most learners are taking matters into their own hands. According to the report, “of these 8.66 million people learning AI, 32.8% are doing so via a structured and supervised learning program, the rest are doing so in an independent manner.” 

Even within structured programs, very few involve colleges or universities. As the report notes, “only 0.2% are learning AI via a credit-bearing program from a higher education institution,” while “the other 99.8% are learning these skills from alternative education providers.” That includes everything from online platforms to employer-led training — programs built for speed, flexibility, and real-world use, rather than degrees.

College programs in AI are growing, but they’re still not reaching enough people. Between 2018 and 2023, enrollment in AI and machine learning programs at U.S. colleges went up nearly 45% each year. Even with that growth, these programs serve only a small slice of learners — most people are still turning to other options.

This disconnect has given rise to a growing field of alternative education providers, many of them operating outside traditional academia. These include online platforms, training startups, and nonprofit initiatives that emphasize faster, more accessible ways to build AI skills. Some focus on project-based learning, others on short-form credentials or industry mentorship, but all are positioning themselves as responses to the institutional lag. 

(ankastudio22/Shutterstock)

Meanwhile, the hiring side isn’t slowing down. The number of AI and machine learning engineer roles is expected to nearly triple by 2027. That puts even more pressure on employers to rethink how they recognize talent and fill roles fast.

Even when people manage to learn AI skills outside traditional systems, that’s only half the battle. The next hurdle is proving it. Most employers still lean on formal degrees when making hiring decisions, even if the work itself no longer requires them. The report points out a telling gap: just 21% of people working in AI roles have a graduate degree, yet 51% of job postings ask for one. 

It’s a disconnect that creates frustration on both ends. Learners walk in with new skills but no credentials to show for it. Employers, meanwhile, are trying to fill roles with signals that no longer reflect how people actually learn. That shift is already being felt in the job market, especially among recent college graduates facing rising unemployment and fewer entry-level roles in AI-heavy industries.

One reason online platforms have taken off is that when formal education can’t move fast enough, people look elsewhere, and they’re finding what they need in places like Coursera and Udemy. 

The report reveals that more than 3.5 million people enrolled in generative AI courses on those two platforms in the first 14 months after ChatGPT launched. The appeal is clear: lower cost, more flexibility, and content that evolves with the tech itself. These platforms weren’t built to replace universities, but for a growing number of learners, they’re the first place to go when it’s time to catch up.

(13_Phunkod/Shutterstock)

The report also points to smaller, community-rooted programs that are filling gaps left by traditional systems. One example is VIAI, a nonprofit offering “pay-what-you-can” AI courses for learners who don’t have the time, money, or access to conventional options. These efforts often serve parents, gig workers, and mid-career professionals. 

To help close the gap, the report recommends that employers provide more practical, on-the-job training. It also calls for more local programs that are easy to join and work for people’s everyday schedules. And it says people need better ways to prove what they’ve learned, even if they didn’t earn a traditional degree. The goal should create a system that values skills and opens more paths for everyone. 

Related Items

TPC25 Preview: Inside the Conference Shaping Frontier AI for Science

Collate Focuses on Metadata Readiness with $10M Series A Funding

Promethium Wants to Make Self-Service Data Work at AI Scale