IBM Survey Shows Gap Between AI Oversight Priorities and Worker Experience

IBM says its new global workforce study found a disconnect between the AI oversight skills leaders prioritize and the skills employees consider important. The gap comes alongside workers’ concerns about eroding abilities and who is responsible when AI goes wrong. In IBM’s announcement of the study, 71% of surveyed chief human resources officers identified supervising, validating and overriding AI outputs as the workforce’s most essential skill, while 29% of employees ranked judgment as important.

The IBM Institute for Business Value study draws on two surveys conducted with Oxford Economics from April through June 2026. One covered 1,500 CHROs and equivalent workforce-strategy executives across 21 geographies and 23 industries; the other covered 8,800 full-time employees in 28 countries. The findings describe respondents’ views and reported organizational outcomes, not conditions independently established across all workplaces. The workplace sample is distinct from a separate IBM-commissioned survey of U.S. educators and parents about schools’ preparedness for AI.

On critical thinking and problem framing, the two groups were closer: 57% of CHROs and 49% of employees selected them among the most important skills for the AI era. At the same time, 60% of employees said they worried about skills erosion. Among that group, three in four said AI had already begun to erode at least some of their skills. IBM said 46% of CHROs also identified skills erosion as a top concern.

The survey also found uncertainty over accountability for AI-assisted decisions. IBM said 43% of employees reported that blame falls on them when something goes wrong with AI. Meanwhile, 41% of CHROs believed employees may not feel safe challenging or overriding AI outputs, and 36% of executives said unclear accountability complicates AI deployment.

IBM reported an association between HR leadership’s formal role and employees’ willingness to question AI. Where the CHRO at least shared responsibility for deciding which decisions remain human-led, 76% of employees said they felt safe questioning or overriding AI recommendations. That figure was 43% where HR had only an advisory role. IBM also said 46% of organizations did not involve the CHRO when AI strategy was defined. The published results do not establish that CHRO involvement caused the difference.

AI-related review and correction work was another issue in the survey. IBM said 80% of CHROs believed adoption creates “invisible” tasks such as validating recommendations, fixing errors, providing context and handling exceptions. Separately, 42% of employees said AI increased their workload or that their work went unrecognized. IBM also reported that 72% of organizations made limited or no use of AI within the HR function.

The companion IBM study page uses a different employee measure, saying 38% prioritized skills needed to evaluate AI outputs, compared with the announcement’s 29% figure for employees ranking judgment as important. The published materials do not explain the difference, so the figures should not be treated as the same measure. IBM’s pages also do not provide the full questionnaires, sampling and weighting procedures, response rates, margins of error or country-level results.