The growing use of artificial intelligence in workplaces is raising concerns about an overlooked consequence of automation: employees may become more dependent on AI while experiencing greater mental fatigue and reduced cognitive engagement.
A study appearing in the November 2026 issue of WORK found that higher dependence on AI was associated with increased cognitive fatigue and lower engagement in mentally demanding work.
The findings come alongside IBM research involving 10,300 respondents, which revealed that 60% of employees worry AI is eroding their professional skills.
Together, the studies highlight a challenge for companies implementing AI: improving productivity without weakening the capabilities employees need to assess, correct and manage automated work.
Researchers Examine the Cognitive Costs of AI
Researchers Umut Uyan and Adil İbin examined responses from 246 academic and administrative employees at public universities who regularly used AI tools.
Their peer-reviewed study, first published online on September 30, used statistical modelling to examine relationships between AI dependence, cognitive fatigue and engagement.
AI dependence was positively associated with cognitive fatigue, with a standardized coefficient of 0.569, and negatively associated with cognitive engagement, with a coefficient of -0.174.
Cognitive fatigue also showed a negative relationship with engagement, while the researchers’ mediation analysis suggested fatigue accounted for part of the association between AI dependence and reduced engagement.
The findings do not establish that AI directly causes cognitive decline. The cross-sectional study relied on self-reported responses from a limited occupational group, and its model explained only 11.4% of the variation in cognitive engagement.
Nevertheless, the research suggests that automating tasks does not necessarily eliminate mental strain. Workers may instead devote effort to monitoring systems, checking results and deciding when intervention is necessary.
IBM Finds a Gap Between AI Skills and Workplace Expectations
A separate IBM Institute for Business Value survey questioned 1,500 chief human resources officers and 8,800 employees globally.
Among its findings, 71% of HR executives identified the ability to supervise, validate and override AI outputs as an essential workforce skill, while only 29% of employees ranked judgment as important.
Of employees concerned about skills erosion, approximately three-quarters believed AI had already begun weakening some of their abilities.
The research also identified a less visible productivity problem.
Some 80% of HR executives believed AI creates additional work involving output verification, corrections and exception handling. Meanwhile, 42% of employees said AI increased their workload or left their contributions unrecognized.
IBM Senior Vice President and Chief Human Resources Officer Nickle LaMoreaux emphasized that human capabilities become increasingly important as automation assumes more routine responsibilities.
Productivity Gains Bring New Responsibilities
Additional research published in Frontiers in Psychology examined 642 hybrid and remote knowledge workers across two survey waves separated by seven weeks.
The study explored how reliance on AI-assisted delegation relates to monitoring concerns, repetitive work-related thoughts and difficulty mentally disconnecting after work.
Its findings add another dimension to AI adoption: delegating tasks to software may change how employees experience responsibility rather than simply reducing their workload.
For founders and product leaders, these findings challenge the assumption that greater automation automatically produces better working conditions.
The implications extend to product design, employee training and organizational accountability.
AI systems that complete tasks efficiently still require people who can evaluate their outputs, recognize errors and exercise independent judgment.
The next measure of workplace AI success may therefore involve more than time saved. It may also depend on whether employees retain the ability, confidence and mental capacity to perform work without relying entirely on automated systems.
