Critical Thinking: The Capability We Skipped

Employers have named analytical thinking their number one skill for years.

One employee in five has had any development in it.

‍

That gap comes from Dale Carnegie's March 2026 research, and it costs more now than it used to. AI output arrives polished whether or not it is right, and polish is the thing people read as a signal of accuracy.

‍

A field experiment with 758 consultants, published in Organization Science in March 2026, found large gains on tasks inside the model's range and a 19 percentage point fall in accuracy on tasks outside it. Nothing in the output marks the boundary.

‍

This is a four-page white paper, a 6-minute read, written for senior people leaders deciding where development money goes next. Written by Graham Dobbin, Director of Change Leadership and Workplace Strategy, Dale Carnegie Australia.

‍

What is critical thinking in the workplace?

Critical thinking is the disciplined work of judging whether a claim holds before acting on it. Where the claim came from, what would have to be true for it to be right, and what else would explain the same evidence.

‍

Dale Carnegie frames it as five phases rather than a single act: problem identification, creative thinking, logical analysis, decision-making, and coordination. The cognitive work in those phases has not changed. The conditions they run inside have. Timelines are compressed, authority is distributed, and a machine now sits in the middle of the reasoning.

‍

The practical distinction is between fluency and judgement. Fluency is the ability to get an answer out of the tool. Judgement is the ability to tell whether the answer is any good. Most organisations have measured the first and assumed the second.

‍

Inside the paper

The jagged frontier.

Where AI lifts performance and where it quietly lowers it, from a 758-consultant field experiment, and why nothing in the output marks the line.

The training gap, measured.

What employees report being developed in, against what employers say they need most. The two have been diverging for years.

Five-Phase Critical Thinking.

Dale Carnegie's model, and what changes at each phase once a machine sits in the middle of the reasoning.

Four moves for leaders.

Practical handles that need no budget, no policy and nobody's permission, with what each one changes.

Why now

‍

AI has made answers cheap. It hands us a starting point dressed as an answer, and the dressing is convincing whether or not the content is. Speed arrived uniformly across the work. Accuracy did not, and the boundary between them is invisible from inside the task.

‍

Analytical thinking has held first place among core skills in the World Economic Forum's Future of Jobs Report across successive editions. The development has not followed. For a business that has bought the tools and cannot tell whether anyone is checking the output, this paper is about the layer underneath that.

‍

Key numbers and ideas

20%

of employees report development in critical thinking, against 50% who report training in technology

‍

19 pts

the fall in accuracy when consultants used AI on tasks outside the model's range

AI hands us a starting point dressed as an answer.

Polish and accuracy have come apart. The polish is the half that stayed.

About the research

The paper draws on Dale Carnegie's March 2026 research into workplace capability development; the field experiment by Dell'Acqua and colleagues published in Organization Science on 11 March 2026 (758 consultants); the World Economic Forum's Future of Jobs Report 2025; AHRI's Quarterly Australian Work Outlook, March 2026 (612 senior decision makers); and Microsoft's 2026 Work Trend Index Australian findings (2,000 Australian workers). The Dale Carnegie paper publishes no sample size or geography, so those figures are quoted as released. Written by Graham Dobbin, Director of Change Leadership and Workplace Strategy, Dale Carnegie Australia. Dale Carnegie global research is led by Robert Coleman, PhD, Director of Research and Thought Leadership.

About Dale Carnegie

Dale Carnegie has been studying what changes how people communicate, lead, and influence at work since 1912. Over 9 million graduates across 90+ countries. In Australia, we work with HR and people leaders at organisations including BHP, Macquarie, GE Healthcare, Schneider Electric, and Veolia to turn capability frameworks into behavioural change you can see in the room.

Download the
Insight Report

Thank you! Your submission has been received!

Oops! Something went wrong while submitting the form.

What percentage of employees are trained in critical thinking?

20% of employees report development in critical thinking, against 50% who report training in technology and 34% in problem solving (Dale Carnegie & Associates, March 2026). The same research found only 30% of leaders describe their organisation's technology and AI integration as truly transformational.

‍

The market named the capability years ago. Analytical thinking ranks first among core skills in the World Economic Forum's Future of Jobs Report, with seven in ten employers calling it essential, and it has held first place across successive editions. The development has not followed the naming.

‍

Is AI making people worse at their jobs?

On some tasks, measurably. The pattern is specific rather than general.

‍

In the 758-consultant field experiment, work sitting inside the model's range came back 12.2% higher in volume and 25.1% faster, with quality up at the same time. On tasks outside that range, consultants working without AI got 84.5% of answers correct. Those working with it got 60%.

‍

The finding that matters most sits alongside those two. Answers produced with AI assistance were rated more coherent and more persuasive whether or not they were correct. The researchers called the boundary a jagged frontier, and the point is that it is invisible from the inside. Speed arrived uniformly across the work. Accuracy did not.

‍

Senior leaders name the same problem when asked. In a BCG survey of 70 C-suite leaders, close to 90% cited overreliance on AI output without stress testing as a concern, while one company in ten had an organisation-wide response to it. Seventy executives is a small base, so read it as a signal from the top of large organisations rather than a measurement of them.

‍

What does the critical thinking gap look like in Australia?

Australian employers name it as a leadership problem. 43% identify problem solving as a critical leadership capability, and 36% report that their own leaders are not fully proficient in it (AHRI Quarterly Australian Work Outlook, March 2026, from 612 senior business decision makers).

‍

Australian workers agree about what will matter. Asked which human skills count most as AI takes on more of the work, they rank critical thinking and quality control of AI output equal first, at 50% each (Microsoft Work Trend Index 2026, Australian findings, from 2,000 Australian workers).

‍

Gartner forecasts that 30% of enterprises will see decision-making quality decline through overreliance on AI by 2030. That is a forecast rather than a measurement, and worth reading as an analyst's view of direction.

‍

Where does judgement enter an AI-mediated decision?

At five points, one for each phase of the model. What changes is not the thinking but the conditions around it.

Problem identification. Whether the problem suits automation at all is a judgement made before any work starts. Introduced too early, AI distorts the goal.
Creative thinking. Generative tools expand ideation and invite early convergence. The discipline is treating machine ideas as inputs and generating human alternatives alongside them.
Logical analysis. Analysis has become an audit. Where did the data come from, and does the output hold across different prompts and versions.
Decision-making. AI compresses the timeline. It does not resolve who carries the consequence, how reversible the call is, or where the risk lands.
Coordination. Humans and algorithms execute together now. Somebody has to draw the line between judgement and automated action, and set what escalates.

‍

How should critical thinking development be measured?

Not by course completions, and not by tool usage. Both measure attendance.

‍

The measures worth having sit in the work. How often AI output moves on without a second pair of eyes. How many corrections are made downstream, and by whom. Whether anyone challenged a recommendation in the last quarter, and what happened when they did. Whether the people using the tools most are the people who have had the most development, or the least.

‍

Anyone attaching a fixed performance uplift to a critical thinking course is going past what the evidence supports. What the evidence does support is that the capability is nameable, teachable and observable in the decisions people make, which is where it should be measured.

‍

More Resources

View all

Insight Report

Rapport Gets the Meeting: It Doesn't Keep the Room

When answers are cheap and everyone can be creative, the advantage sits with the people who can read a client's business, ask the question nobody has asked, and make a case worth defending.

Insight Report

AI in the Workplace: What Actually Changed

Almost nine in ten organisations now use it. Four in five people using it say they are more productive. Only six percent can point to it in the profit line.

Your Cart

Critical Thinking: The Capability We Skipped

Employers have named analytical thinking their number one skill for years.