IBM Survey Finds 60% of Employees Worry AI Is Eroding Skills
The study also surfaces hidden validation work and accountability gaps, but its self-reported results do not establish that AI caused the reported outcomes.
Edited by Tyronne Panaino
IBM published a workforce study on September 21 reporting that 60% of surveyed employees worry AI is eroding their skills, with critical thinking cited most often as declining. The IBM Institute for Business Value worked with Oxford Economics on two surveys conducted from April through June 2026: one covered 1,500 CHROs and equivalent workforce executives, and the other covered 8,800 full-time employees.
The findings matter because they measure more than AI adoption. IBM's newsroom summary and full study page describe a gap between the evaluative skills leaders expect, the work employees say AI adds and the authority workers feel they have when an AI-supported decision goes wrong.
Leaders and employees rank AI oversight differently
IBM reports that 71% of executives prioritize the ability to supervise, validate or override AI output, compared with 38% of employees. The same study says only 26% of organizations clearly divide work into human-led, AI-assisted and AI-executed activities.
Those figures point to a design problem rather than a simple training deficit. An employee can learn to use an AI tool without knowing which output must be checked, who may override it or where responsibility moves when automation acts. The study also says 52% of employees report that their assigned tasks changed during the previous year as AI reshaped daily responsibilities.
IBM links clearer judgment roles with employee confidence: 62% of CHROs in organizations where judgment is built into work report rising confidence in AI-enabled decisions, while 57% report declining confidence where it is not. That is an association reported by executives, not evidence that one governance change alone caused the difference.
AI can create checking work that is hard to see
Eighty percent of surveyed CHROs say AI adoption creates hidden work such as validating recommendations, correcting mistakes, supplying context and managing exceptions. On the employee side, 42% say AI either increases their work or leaves that work unrecognized.
Accountability is another pressure point. IBM reports that 43% of employees say they are blamed when something goes wrong with AI, while 41% of CHROs believe employees may not feel safe challenging or overriding an AI output. The study also finds that 72% of organizations make limited or no use of AI inside HR itself, even as HR leaders are expected to shape the wider workforce transition.
Taken together, the results suggest that measuring time saved can miss the effort shifted into review, exception handling and responsibility. Teams evaluating an AI workflow need to count that work explicitly and define the authority to stop or correct the system.
What the survey can and cannot establish
The executive survey covered 21 geographies and 23 industries, while the employee survey spanned 28 countries. That breadth is useful, but the fetched pages do not provide an independent replication or establish that the samples represent every national workforce. The results are also self-reported and come from IBM's own research organization.
IBM frames higher-performing respondents as Thinking Organizations and compares their reported capabilities and outcomes. Those comparisons can identify patterns worth testing, but they do not by themselves prove that a specific AI-work design caused revenue, confidence or quality changes. A stronger next checkpoint would be longitudinal evidence that tracks the same roles before and after workflow redesign, including time spent validating AI and the quality of override decisions.
Status
Learning. IBM's official study pages support the sample, percentages and research framing. Internal confidence is medium because the evidence is first-party survey research, the outcomes are self-reported and causal effects have not been independently established here.
Sources
- IBM Newsroom — 2026 CHRO workforce study
- IBM Institute for Business Value — Designing the Thinking Organization
Update note: Last reviewed 2026-09-24. We will revise this post if IBM publishes fuller sampling detail, longitudinal results or an independent replication.
Sources
Drafted with AI assistance from source briefs; reviewed for citation completeness and label accuracy.