---
title: Using AI Can Weaken the Judgment Your Company Needs Most
description: Leaders now rank human judgment as the top skill in an AI workplace. Their workers fear it is eroding, and independent research backs the worry.
author: Darie Nani (Editor-in-Chief)
date: 2026-09-21T13:15:10.777Z
updated: 2026-09-21T13:15:10.797Z
canonical: https://www.sovereignmagazine.com/article/ai-weakens-workforce-judgment-critical-thinking
image: https://cdn.nanimediahouse.com/pexels-professional-woman-in-office-setting-studying-documents-with-7648004.jpg
categories: Artificial Intelligence
content_type: Analysis
region: Global
publication: Sovereign Magazine
schema_type: Article
---

Chief human resources officers now name one skill above all others as essential to their workforce: the ability to supervise, validate and override what AI produces. Their employees mostly do not share that view, and many fear the skill is already slipping away.

In a study published September 21 by the IBM Institute for Business Value with Oxford Economics, 71% of the 1,500 CHROs surveyed ranked that capacity for judgment as the workforce's most essential skill. Among the 8,800 employees surveyed, only 29% rated judgment as important. Sixty percent said they worry AI is eroding their skills, and the one they named most often as declining was critical thinking.

In the CHROs' broader ranking, 57% placed critical thinking among the most important workforce capabilities and 48% named human judgment.

Nickle LaMoreaux, senior vice president and chief human resources officer at IBM, tied the shift to what companies are automating first. "AI is changing not only how work gets done, but where people can contribute the greatest value," LaMoreaux said. "As AI takes on more routine and process-driven tasks, uniquely human capabilities become even more important."

The routine, process-driven work companies hand to AI first is often the same repetition through which people build and keep their judgment sharp. Remove the practice, and the skill can fade with it.

## Two Studies Link Heavy AI Use to Weaker Thinking

A 2025 study by researchers at Microsoft Research and Carnegie Mellon University, presented at the CHI conference, surveyed 319 knowledge workers who supplied 936 first-hand examples of using generative AI at work. It found that [a worker's confidence in the AI predicts how much critical thinking they do](https://www.microsoft.com/en-us/research/publication/the-impact-of-generative-ai-on-critical-thinking-self-reported-reductions-in-cognitive-effort-and-confidence-effects-from-a-survey-of-knowledge-workers/): the more they trusted the tool, the less critical thinking they reported applying. The more confident they were in their own ability, the more they applied.

A separate 2025 study from the MIT Media Lab, titled "Your Brain on ChatGPT," measured what happens in the moment of use. Researchers used EEG to track the brain activity of 54 people writing essays across four sessions. Participants who [leaned on a large language model showed the weakest and least connected brain activity](https://arxiv.org/abs/2506.08872), while those who wrote unaided showed the strongest. The AI users also struggled to accurately quote work they had just produced and felt the least ownership of it. The authors call the effect "cognitive debt."

> "Fluency without judgment simply helps an organization make mistakes faster."
> — Dr. Amit Das, chief human resources officer, Bennett Coleman & Co.

## Defining Where People Stay in Charge Cuts the Risk

Organizations that clearly define each workflow as human-led, AI-assisted or AI-executed report 18% lower risk and 20% better quality than those that leave the boundary vague. The gain comes from deciding in advance where a person validates the output and where the machine runs unsupervised, rather than letting the line blur task by task.

The study also found that 46% of organizations do not involve the CHRO when AI strategy is decided, which leaves the people responsible for the workforce's skills outside the room where the tools that shape those skills are chosen. The workers in the Microsoft and Carnegie Mellon study who stayed confident in their own ability kept thinking critically, so the practical task is to build workflows that require a person to check, question and override, not just accept.

A leader can set that boundary in advance for every workflow; keep a person in the loop on the outputs that carry risk, so validation stays a habit rather than a formality; and bring HR into AI strategy, so the cost to skills is weighed when tools are chosen rather than discovered once judgment is already thinner.

## FAQ

**Q: What does "cognitive debt" mean?**
It is the term MIT Media Lab researchers use for the mental engagement a person gives up when they lean on AI to do their thinking. In their study, people who wrote with a large language model showed weaker, less connected brain activity than those who worked unaided, and they struggled to recall or accurately quote work they had just produced. The worry is that the effect builds up the more often the thinking is handed over.

**Q: Can AI also improve critical thinking?**
The Microsoft and Carnegie Mellon research offers a qualifier: workers who were more confident in their own ability applied more critical thinking, not less. What matters is whether the person stays engaged in checking and questioning the output, which points to how the tool is used rather than whether it is used at all.

**Q: Which tasks are safe to automate, and which should stay human-checked?**
The approach the study points to is to decide in advance for each task rather than case by case. Routine, low-stakes and process-driven work is the easiest to hand to AI; the outputs where a wrong answer is costly are the ones to keep under human validation, because that is where lost judgment does the most damage.