New IBM CHRO Study: AI Puts Critical Thinking at the Center of Workforce Priorities

You have probably sat in a town hall where the chief executive promised that AI would augment, not replace, your team's expertise. Then you watched colleagues hand prompt-writing to the platform and stop asking whether the output made sense. The IBM Institute for Business Value has now quantified that unease: 60% of employees worry AI is eroding their skills, and critical thinking tops the list of capabilities they believe are declining.

The finding arrives at a moment when organisations treat AI adoption as a technology deployment rather than a capability shift. Nearly half of the companies surveyed do not involve the Chief Human Resource Officer when AI strategy is being defined, which means the people responsible for workforce development learn about new tools after procurement signs the contract. The result is a gap between what CHROs say AI demands and what employees prioritise once the system goes live.

What CHROs want versus what employees need

71% of CHROs identify the ability to supervise, validate and override AI outputs as the most important skill in an AI-augmented workplace. That framing treats the human as quality-control gate rather than decision-maker, and it explains why employees feel their judgement is atrophying. When the job becomes checking a machine's homework instead of forming an argument from first principles, critical thinking moves from daily practice to occasional override.

The gap is not a training deficit. It is a design problem. If your learning-and-development programme teaches people to spot hallucinations but not to build the mental model that would let them write the brief without a prompt, you are preparing them to be good reviewers in a world that still rewards good thinkers. The CHRO's supervision skill and the employee's critical-thinking skill are related, but they sit at different altitudes: one assumes the machine produces the draft, the other assumes the human owns the reasoning.

The governance vacuum

When half of organisations exclude the CHRO from AI strategy conversations, they are making a choice about what AI is for. If the technology is a back-office efficiency play—faster contract summaries, automated expense approvals—then keeping HR out of the room is merely short-sighted. If the technology reshapes how analysts form hypotheses, how designers iterate and how managers allocate attention, then excluding the CHRO is a governance failure. Workforce capability is not a downstream consequence of AI; it is the mechanism through which AI creates or destroys value.

The absence of the CHRO also means that the people defining use cases do not carry accountability for skill erosion. A product team that builds an AI feature to reduce time-to-insight will celebrate cycle-time gains and ignore the fact that junior staff no longer learn to read a data table. A finance leader who automates variance analysis will not notice that analysts stop understanding why two numbers diverge. The CHRO is the only executive whose performance metrics degrade when capability declines, which makes their exclusion from strategy a design flaw, not an oversight.

What this means for L&D

If 60% of your people believe AI is eroding their skills, the fix is not a critical-thinking workshop. The fix is changing how you introduce AI tools in the first place. That means three moves:

• Teach the reasoning before you teach the tool, so that employees understand what good looks like and can judge whether the machine's output meets that standard, rather than learning to accept the first plausible answer.
• Design AI use cases that require human judgement at the problem-definition stage, not just the validation stage, so that critical thinking remains a daily practice instead of an occasional audit.
• Involve the CHRO when you scope the AI strategy, because the person responsible for capability development needs to see the use case before the platform goes live, not after adoption metrics are already being tracked.

The IBM study does not say that AI makes people less capable. It says that 60% of employees feel less capable, and that perception gap is a leading indicator. If your workforce believes the tools are replacing their judgement, they will stop exercising it, and the atrophy becomes real. The CHRO's job is to prevent that loop from closing, but they cannot do it from outside the room where AI strategy is written.

Sources:

• IBM Newsroom, 21 September 2026