“Using that new tool will degrade the skill levels for students today and for all future generations!” This was the concern raised not in 2026, but in the mid-1970s as the rapid adoption of the handheld calculator changed classrooms and workplaces. While the debate and research on how new technologies impact humans continue, we now find ourselves facing the rapid adoption of artificial intelligence with unknown consequences for education and human development.
Much of the conversation about AI and work has centered on displacement—which jobs will vanish, which will emerge. A quieter crisis is unfolding alongside it: deskilling. Workers are at risk of losing the very skills that make them employable, even as they keep their jobs.
AI’s Impact on How We Think
A study by researchers at Microsoft and Carnegie Mellon University surveyed 319 knowledge workers across nearly 1K real-world AI-assisted tasks. The more workers trusted AI’s capabilities, the less they engaged in critical thinking about its outputs. Workers with access to generative AI tools also produced a less diverse set of outcomes for the same task compared to those without. A separate study at SBS Swiss Business School found a significant negative correlation between AI usage and critical thinking scores, with cognitive offloading—the tendency to delegate mental work to a tool—acting as the primary mechanism. Younger participants, ages 17 to 25, showed the highest AI dependence and the lowest critical thinking scores.
Researchers at Aalto University in Finland examined an accounting firm where reliance on automation fostered complacency and eroded staff awareness, competence, and the ability to assess outputs. An MIT Media Lab study reported that excessive reliance on AI-driven solutions contributes to cognitive atrophy. The pattern is consistent across industries and methodologies: AI automates the processes through which people develop and maintain their skills—the tasks and the learning that used to come with them.
Of course, every transformative technology triggers both excitement and anxiety given the change it brings to work, education, and society. When calculators entered classrooms, critics warned they would destroy arithmetic ability. After countless research studies, the evidence generally showed that students with sustained calculator access used a wider range of problem-solving approaches. However, studies also documented significant erosion of mental calculation skills when calculators were introduced too early and basic math skills were not developed. The lesson from calculators: the design of the learning environment around the tool determines whether skills atrophy or deepen.
AI presents this same fork in the road, at far greater scale. Calculators affected arithmetic. Search engines affected memory and research habits. AI affects reasoning, writing, analysis, coding, and decision-making simultaneously—and delivers outputs that look polished and authoritative even when they’re wrong. At the same time, more than 90% of business leaders expect greater AI integration in their organizations in 2026.
What This Means for Learning and Development
People have shown resilience in adapting to new technologies. While many advancements have helped reduce physical demands, such as automobiles, electricity, and computers, the impact of AI is quite different as it aims squarely at human cognition. This has implications for employers, educators, and policymakers.
Many employers recognize the potential impact of AI and the need for retooling or upskilling their workforce. BCG’s 2025 global study of C-suite executives found that only 5% of companies are generating substantial value from AI—and that 70% of what separates them from laggards comes from investing in people, not technology. Future-built companies upskill more than 50% of their workforce on AI, compared with 20% at lagging firms. Companies that push AI into every workflow under the banner of efficiency risk producing employees who appear productive on paper but cannot perform without digital scaffolding. Some level of employer-sponsored retooling will be critical to success.
In education, the challenge is structural. AI is compressing middle-skill knowledge work and eliminating the entry-level roles that once served as on-ramps for college-educated workers, as AEI’s Brent Orrell argues in a recent piece for the National Association of State Boards of Education. Young people best positioned to thrive will combine legacy technical skills with AI literacy and capacities like critical thinking, communication, and ethical reasoning. Most secondary school curricula aren’t designed for any of those.
Policymakers are concerned about the impact of AI on the economy and generally agree that retraining will be needed. Yet, a recent Upjohn Institute report found that current workforce programs are underfunded, backward-looking, and poorly aligned to local demand, even as retraining ranks as the top policy preference across party lines in both the U.S. and Canada. The policy responses to displacement—Workforce Pell, apprenticeship expansion, and short-term credentialing—are necessary but insufficient if they ignore the deskilling dynamic. The IMF’s January 2026 report makes clear that policy choices made now will determine whether workers and firms are prepared for what’s ahead.
Employee training programs and school curricula should explicitly build cognitive resilience, the capacity to reason independently before and after consulting AI. The Microsoft-CMU study’s most actionable finding is that workers who had less confidence in AI’s ability engaged in more critical thinking, suggesting that training environments benefit from structured friction where workers reason on their own first. Employers should track whether workers are building transferable skills or merely tool dependence. Policymakers may wish to measure human development in a different way to help ensure that education and retooling are focused on gaining durable competencies rather than narrow proficiency.
The deskilling default is already here. Without deliberate intervention from employers, educators, and policymakers, AI adoption will hollow out the cognitive capabilities it is designed to augment.
Richard R. Smith is a professor at Johns Hopkins University, where he also serves as the faculty director for the Human Capital Development Lab. Arafat Kabir is an advisor at the Johns Hopkins Human Capital Development Lab and a contributing writer at Forbes, examining how AI is transforming work and society.