Rashmeet Kaur Generative AI is rapidly reshaping how employees approach learning, with a growing number eager to use the technology for upskilling. Yet a knowledge gap persists — and it’s proving to be a bottleneck for employers.
A sweeping majority of U.S. workers — 86%, according to a survey by EdAssist — are motivated to use AI to learn new skills. But many are running into the same wall: they don’t know how to use it well enough to learn anything meaningful.
The problem isn’t access. It’s understanding. Employees often don’t know how to frame effective prompts or critically evaluate AI-generated outputs. That makes even the most powerful tools underwhelming in practice.
“AI can help people learn faster — but only if they know how to interact with it,” says Diane Bartoli, SVP at EdAssist. “And right now, most don’t.”
Rather than prescribe a one-size-fits-all approach, companies are tailoring AI training by role and industry. Soft skills — once viewed as innate — are getting particular attention. Communication, critical thinking, and leadership are now considered essential, teachable proficiencies. “We used to assume people either had these skills or didn’t. That’s no longer viable,” Bartoli adds.
The gap is widest among Gen Z workers. They’re the most enthusiastic adopters of AI — 43% want to use it to improve work-life balance compared to 35% of all age groups, but also the most at risk. Many early-career workers haven’t yet developed the judgment to vet AI’s suggestions. They’re also looking for more meaningful work — but often lack the foundational experience to judge AI’s outputs, making targeted support even more critical.
In fields like banking, where junior analysts used to learn by doing, AI now automates many of those formative tasks. Without hands-on exposure, young professionals risk becoming over-reliant on machine suggestions without the judgment to assess them.
The bottom line: AI won’t replace learning, but it’s rewriting how it happens. Companies that can teach workers how to learn again, will have the edge.
