The AI Skills Gap: What Employers Actually Want
The AI skills gap is less about workers falling behind a technology and more about employers wanting something specific and struggling to find it in the people who apply. Knowing what that specific thing is makes it easier to close the gap on your end.
The World Economic Forum's Future of Jobs Report 2025, based on data from more than 1,000 employers, found that 63 percent of them name skill gaps as the main barrier to transforming their business (WEF). No other barrier in the survey ranked higher.
What the skills gap actually means
"Skills gap" sounds abstract, but employer surveys get specific about where it bites. In ManpowerGroup's 2026 Talent Shortage Survey of about 39,000 employers in 41 countries, AI model and application development topped the list of hardest skills to find (cited by 20 percent of employers), with AI literacy tied with engineering for second at 19 percent, ahead of sales and marketing and manufacturing (Staffing Industry Analysts). So a large part of the gap is basic AI literacy: knowing how to use AI tools competently, judge their output, and work alongside them, at every level of a company and in non-technical roles too. A shortage of elite machine learning engineers is only one piece.
The WEF report also found that employers expect 39 percent of workers' core skills to change by 2030. If the global workforce were 100 people, 59 would need reskilling or upskilling by then, and 11 of them are unlikely to get it. Skills need to change, and a real share of the workers who need retraining won't receive it through an employer. Some of the responsibility falls on the individual.
What employers say they're doing about it
Most employers say they plan to respond with training. In the WEF survey, 85 percent plan to prioritize upskilling their workforce, and 77 percent plan to upskill workers in response to AI. The same survey also found 41 percent plan to reduce their workforce where AI automates tasks, so training and cuts are both on the table.
If you're currently employed, that's a mixed but useful signal: many companies would rather train the people they have. "Planning to upskill" and "delivering training that works" are different things, though, and the reskilling numbers above show the plan doesn't reach everyone. If your employer hasn't offered AI training yet, that's common. It's also a reason to look for training yourself instead of waiting.
The gap looks different by field
Hiring data shows the gap isn't spread evenly. Indeed's Hiring Lab found that by the end of 2025, U.S. job postings mentioning AI were 134 percent above their February 2020 level, while total postings were only about 6 percent above it (Indeed Hiring Lab). Mentions are most common in data and analytics roles (about 45 percent of postings), but they're growing in marketing, HR, finance and project management too.
Employers are also asking for AI literacy in jobs that have nothing to do with technology. Employment law firm Fisher Phillips, writing for hiring employers, says postings that list "AI literacy" as a requirement tripled in 2025, and notes there is still no shared definition of what the term means (Fisher Phillips).
That matters if you're early in your career or changing fields. The old assumption, that you'd pick up the tech tools later on the job, holds less well than it used to. More employers expect a baseline of AI comfort on day one, even in roles outside tech.
Closing the gap: what helps
The employer data points to a short, practical list. None of it is a mysterious technical curriculum.
- Basic AI literacy first. Since AI literacy is near the top of the hard-to-find list, being competent with common AI tools (knowing what they're good at, what they get wrong, and how to check their work) puts you ahead of many applicants.
- Analytical thinking and judgment. The WEF survey ranks analytical thinking as the most sought-after core skill, with seven in ten companies calling it essential. This is the skill of deciding what's worth doing and checking whether an output, AI-generated or not, is correct.
- Resilience, flexibility and agility. WEF ranks this group second. Employers expect tools and workflows to keep changing, and they want people who adjust.
- Domain expertise plus AI fluency. Employers rarely say they can't find anyone who can use a chatbot. The harder find is someone who understands the field and can also direct AI tools inside it. Depth in your field is still the foundation. AI comfort sits on top.
None of these require a computer science degree. Free and low-cost courses on using AI tools (as opposed to building them) cover much of the literacy piece. Judgment and adaptability are harder to shortcut. They mostly come from practice, including enough regular use of AI tools to get a feel for where they help and where they mislead.
Why this connects to job security
The skills gap also previews where AI exposure lands inside a role. Tasks built around routine information processing are the ones current AI tools handle best, while work that needs physical presence or personal accountability is harder to hand off. Closing your own piece of the gap, so you're the person who directs and checks AI work rather than only doing what a tool now does cheaply, is one of the more concrete things you can do about that exposure. Our breakdown of the most AI-proof jobs covers which categories carry the least of that risk to begin with.
Where you actually stand
Survey data is useful, but it stays abstract until you check it against your own job. The tasks in your actual week, more than your title, decide how exposed you are and how much the skills gap applies to you.
Take the free "How AI-Proof Is Your Job?" assessment for a private, anonymous read on where your role stands, based on the tasks you actually do.
Keep the skills that keep you employed
The tasks you still do by hand, without checking a tool first, are part of what keeps you valuable. The free 5-Day AI Reset is a five-email course built around that idea: Day 2 has you take one task back and do it unassisted. One small change per day.
Frequently asked questions
What is the AI skills gap?
It's the mismatch between the AI-related skills employers say they need and the skills available in the workforce. In the World Economic Forum's Future of Jobs Report 2025, 63 percent of employers named skill gaps as the main barrier to transforming their business, and ManpowerGroup's 2026 survey ranked AI skills as the hardest to find.
What skills do employers want most?
In the WEF survey, analytical thinking is the most sought-after core skill, with seven in ten employers calling it essential, followed by resilience, flexibility and agility. On the technical side, ManpowerGroup's 2026 survey put AI model and application development first on the hard-to-find list, with AI literacy tied for second.
Is the AI skills gap mostly a technical problem?
Only partly. A large share of it is basic AI literacy: knowing how to use AI tools competently and judge their output. That makes it more solvable for most workers than a shortage of specialized engineers would be.
Will my employer train me on AI skills?
Maybe, but it isn't guaranteed. The WEF survey found 77 percent of employers plan to upskill workers in response to AI, but the same report estimates that of every 100 workers, 11 who need reskilling by 2030 are unlikely to get it. It's worth pursuing training on your own instead of waiting.
How do I start closing the AI skills gap myself?
Build basic comfort with the AI tools common in your field, keep sharpening the judgment and analytical skills employers rank highest, and pair both with the domain expertise you already have. Employers describe wanting depth in a field plus AI fluency, not AI fluency on its own.
Take the free "How AI-Proof Is Your Job?" assessment and see where your own tasks stand, free and anonymous.