Gen Z Workers Struggle to Apply AI-Assisted Learning on the Job, Survey Finds

Despite near-universal adoption of artificial intelligence for professional development among younger employees, a significant majority report difficulty translating chatbot-generated knowledge into practical workplace performance. A new survey of over 5,000 professionals across nine countries reveals that while 92% of Gen Z workers used AI to learn career-related skills in the past year, 57% struggled to apply that information in real-world scenarios.

This application gap is notably wider than among older cohorts. Only 40% of Gen X respondents reported similar difficulties putting new learning into practice. The data suggests that high-frequency AI use does not automatically correlate with job readiness or skill mastery, raising questions about how organizations structure training programs and support early-career talent.

The disconnect appears tied to the nature of AI interaction itself. Large language models excel at providing instant explanations, generating formulas, or summarizing concepts, but they lack the contextual feedback loops inherent in human mentorship. When an employee asks an AI for a spreadsheet function, the tool provides the syntax without ensuring the user understands the underlying logic. If a manager later requests a modification or challenges the output, the worker may be unable to adapt because their learning was procedural rather than conceptual.

This phenomenon mirrors broader trends in digital upskilling. Across all age groups surveyed, 76% of working adults said they spend more time learning now than two years ago, yet 52% feel this increased investment has not yielded better results. For employers, many of whom fund these development efforts—51% of respondents received partial or full employer sponsorship—the return on investment remains uncertain when learners cannot operationalize what they have studied.

Industry experts emphasize that AI functions best as a content generator rather than a complete pedagogical system. Josh Crossick, chief product officer at Preply, which conducted the research, noted that effective skill acquisition requires three distinct phases: knowledge acquisition, deliberate practice, and confidence building through real-time feedback. AI can accelerate the first phase but often fails to support the latter two.

Human coaching fills those gaps by introducing productive friction. In live interactions, mentors read nonverbal cues, adjust explanations based on confusion, and force learners to recover from mistakes rather than simply regenerating a cleaner answer. This social dimension of learning builds resilience and situational awareness that static AI responses cannot replicate. Todd Hughes, a language trainer at Rosetta Stone, described AI as a vast library useful for sourcing materials but insufficient for developing communicative competence without interpersonal rehearsal.

Remote and hybrid work environments may be amplifying these deficits. Iain Smith, head of behavioral science at Sunny Workplace, pointed out that junior staff increasingly miss informal learning opportunities that historically occurred through proximity to experienced colleagues. Observing how veterans handle stress, navigate office politics, or troubleshoot problems provided tacit knowledge that formal training rarely covers. When onboarding happens primarily through screens, new hires lose access to this apprenticeship model and may over-rely on AI as a substitute for institutional wisdom.

The survey also found that 71% of Gen Z respondents felt busy learning without improving, compared to 56% of Gen X. While the study did not isolate AI as the sole cause, the correlation between heavy AI reliance and perceived stagnation suggests that unstructured self-directed learning may create an illusion of progress. Without mechanisms to validate understanding or integrate new knowledge into team workflows, employees accumulate information without developing capability.

For organizations, the implication is that deploying AI learning tools is necessary but insufficient. Companies must pair technological access with structured coaching, peer feedback systems, and opportunities for applied practice. Crossick emphasized that understanding “the human side of the system” remains critical; tools provide information, but experienced colleagues provide the judgment needed to use it effectively.

Preply released the survey findings Thursday, highlighting the need for employers to reassess how they measure learning outcomes beyond completion rates or hours logged.