The AI Skills Trap: 55% of Workers Use AI, Only 33% Got Any Training

The Numbers Behind the Gap

The headline finding is a mismatch between how much workers use AI and how little they are trained to. According to The Conference Board, in a study of nearly 1,300 workers globally titled “Skilling for AI,” 55% of workers use generative AI or AI agents on a daily or weekly basis. Yet only 33% participated in any employer-provided AI training in the past six months, and 28% report their organization provides no AI training at all.

The resource picture is equally thin. The same study, published in June 2026, found that fewer than half of workers believe their organizations give them sufficient time (48%) or adequate tools, access, and resources (48%) to develop AI skills. In other words, a majority are already using AI in their work while a majority also lack the time, tools, or training their employers should be providing.

The scale of the coming disruption makes the gap urgent. The World Economic Forum’s Future of Jobs analysis projects that employers expect 39% of workers’ core skills to change by 2030, and that job disruption will equal roughly 22% of jobs — with 170 million new roles created and 92 million displaced, for a net 78 million new jobs by 2030, per the WEF’s own summary. When a third of the workforce’s skills are set to churn and most workers say they lack the resources to keep up, the training deficit stops being a soft benefit and becomes a strategic risk.

Skilling for Today’s AI, Not Tomorrow’s Jobs

The most important finding is not a number but a diagnosis. The Conference Board concludes that most organizations are focused on helping employees grow in their current roles rather than preparing them for the jobs of the future. Companies are investing in AI literacy and current-role upskilling — teaching people to use today’s tools in today’s jobs — while few are preparing workers for the reskilling that AI-driven transformation will actually require.

This is a subtle but decisive distinction. Teaching an analyst to use an AI assistant makes them more productive in the job they have. It does nothing to prepare them for a future in which the analyst role itself is reshaped, merged, or partly automated. The training that feels responsible today — “we rolled out AI-literacy modules” — can be aimed at a target that is already moving.

The corroborating evidence is consistent. As summarized in The Conference Board’s release via PR Newswire, a separate reading of the same disruption shows 42% of employees expecting their role to change significantly within a year while only 17% use AI frequently — an adoption-and-anticipation gap that mirrors the training gap. Workers sense the change is coming faster than their skills, and their employers’ programs, are moving.

The Analysis: A Shared Failure With Uneven Consequences

Read carefully, the data describes a shared failure of employers and workers to prepare for a transition both know is coming — but the consequences fall unevenly.

For employers, the risk is strategic. A workforce that uses AI without training uses it badly, insecurely, and inconsistently — and a workforce trained only for today’s tools will be stranded when roles change. The Conference Board’s framing implies that many current training budgets, however well-intentioned, are being spent on the wrong horizon.

For workers, the consequences are personal and unequal. Those in organizations that provide time, tools, and forward-looking training will compound their advantage; those in the 28% of workplaces offering no AI training at all are being left to self-teach or fall behind. This is how skills gaps become inequality: the same technology that could level the field instead widens it, because access to good training is unevenly distributed.

The honest read is that the AI skills problem is not primarily a technology problem or even a training-volume problem — it is a targeting problem. Organizations are training, but often for the wrong future. And the workers most exposed are precisely those with the least support to correct for it themselves. The number that should worry leaders is not the 55% using AI; it is the gap between that figure and the 33% being trained — and the fact that even much of that 33% is aimed at today, not tomorrow.

What Workers and Employers Should Do

The gap is real, but it is also actionable. For workers and organizations willing to aim training at the right target, the study points to concrete moves.

1. Workers: treat AI upskilling as a personal responsibility, not a perk you wait for

With 28% of employers offering no AI training and only 48% providing sufficient time, waiting for your employer is a losing strategy. Build a deliberate self-learning habit around the AI tools in your field, and document what you learn — the skills compound, and the record helps you switch employers if yours won’t invest.

2. Employers: split your training budget between “today’s tools” and “tomorrow’s roles”

Most programs teach current-tool literacy. Deliberately carve out a portion for reskilling toward the roles your organization will need in three years, not just the tasks it runs today. The Conference Board’s core warning is that skilling only for the present leaves both the company and its people exposed.

3. Both: close the time-and-tools gap before adding more content

With only 48% of workers reporting enough time or resources, more training modules will not help if people cannot actually use them. Protect learning time and provide real tool access first — otherwise “training offered” never becomes “skills gained.”

4. Workers: build the durable skills AI does not replace

As roles churn, the resilient skills are judgment, problem framing, communication, and the ability to direct AI rather than merely operate it. Invest in the human capabilities that survive automation alongside the tool-specific skills that will keep changing.

Where This Fits in 2026’s Careers Landscape

The Conference Board’s study is a snapshot of a workforce caught mid-transition: using AI heavily, trained for it lightly, and prepared for the future least of all. The deeper lesson is about aim, not volume. Organizations are not failing to train — many are training earnestly — but they are training for the AI and the jobs of today while the WEF’s own projections point to roughly a third of core skills changing and tens of millions of roles being displaced and created by 2030. The workers and companies that thrive will be those who close two gaps at once: the visible one between using AI and being trained for it, and the invisible one between training for the present and preparing for the future. In a labor market this fluid, the safest career strategy is to assume your own upskilling is your job — and to aim it, deliberately, at the roles that do not exist yet.

Frequently Asked Questions

Sources & Further Reading

• Report: Most Organizations Are Preparing Workers for Today’s AI, Not Tomorrow’s Jobs — The Conference Board
• Report: Most Organizations Are Preparing Workers for Today’s AI, Not Tomorrow’s Jobs — PR Newswire / The Conference Board
• The Future of Jobs Report 2025: Skills Outlook — World Economic Forum
• Future of Jobs Report 2025: 78 Million New Job Opportunities by 2030 — World Economic Forum