AI can make us faster, more productive, and seemingly more capable. But if convenience slowly becomes dependency, what happens to the critical thinking, judgement, and creativity we once developed by doing the hard work ourselves?
I believe AI can weaken those skills when it replaces the hard work that builds judgement. The risk is greatest for people early in their careers who have not yet built the experience to judge whether AI is right. But the answer is not to ban AI. It is to keep valuing the tasks that teach us to think.
Not long ago, I wrote something on my personal Facebook page about AI.
It wasn’t a piece of HR thought leadership or a prediction about the future of work. It was simply a personal reflection on something I was finding increasingly uncomfortable: how technology can now imitate our appearance, our voices, our words, our emotions, and even our creativity.
As someone who has always loved words and taken pride in being able to write, I found myself wondering what happens when something I have spent a lifetime developing can suddenly be replicated in seconds.
I use AI regularly. I think it is an extraordinary tool. It helps me research, organise my thoughts, and sometimes find a better way of expressing something I was already trying to say. I don’t want to stop using it.
But increasingly, I find myself asking a different question.
What happens when convenience slowly becomes dependency?
Does relying on AI weaken critical thinking?
It can. When AI replaces the hard work of thinking it removes the very process that teaches us how to think. Research from Microsoft and Carnegie Mellon University backs this up. Their 2025 survey of 319 knowledge workers found that higher confidence in generative AI was associated with less critical thinking.
I have spent many decades working in HR. Long before AI tools existed, I had to learn how to investigate a problem, weigh up conflicting evidence, formulate an argument, write a difficult letter, advise a manager, and sometimes tell a senior leader something they didn’t particularly want to hear.
I got things wrong (I still do), I had to rewrite and rethink things, I stared at blank pages, and I frequently changed my mind. And all of that mattered because going through those processes taught me how to think.
That makes my relationship with AI very different from that of someone entering the workplace today.
When I ask AI for information or to review something I have written, I have enough experience to decide whether what it is telling me is valid and appropriate. I can spot when something doesn’t sound right, when an answer lacks context, or when even if the advice is technically correct, it would be completely wrong or inappropriate in the context of the real-world scenario I am working with.
The research supports this. The Microsoft and Carnegie Mellon study found that people with higher confidence in their own abilities applied more critical thinking to AI output. Experience gives us the confidence to question what AI tells us.
But what happens when someone hasn’t developed that judgement yet?
Why are early-career employees most at risk?
Early-career employees are most at risk because they have not yet built the experience to know when AI is wrong. A 2025 study of 666 people published in the journal Societies found that younger participants relied more on AI tools and scored lower on critical thinking than older participants.
Imagine being 21 or 22 and entering your first professional role today. AI can generate a difficult email, analyse information, and even generate a presentation.
And the temptation to use it is very high. It saves you time, it feels safer.
My concern is that it is often the difficult, frustrating, and time-consuming tasks we need to perform that are actually the things that help us develop judgement, confidence, and critical thinking.
If the first response to every unfamiliar problem becomes “I’ll ask AI”, at what point do we stop developing the ability to work out a problem for ourselves?
There is an important difference between using AI to accelerate your thinking and using it to avoid thinking.
| Task | Using AI to accelerate your thinking | Using AI to avoid thinking |
|---|---|---|
| A difficult email | You decide what needs saying and use AI to find a better way of expressing it. | AI writes it and you send it without judging whether it is right for the situation. |
| Research | AI helps you research and organise your thoughts. | AI's answer becomes your conclusion. |
| Analysis | You check what AI tells you against your own experience and the real-world context. | You accept the output without knowing whether it is any good. |
| A first draft | You produce the first version and learn from the review. | AI produces the first version and the learning experience is lost. |
Are we removing the “apprenticeship of work”?
The apprenticeship of work is how junior employees learn by doing basic work that someone more experienced then reviews and explains.
This is where I think we need to pay attention. For years, junior employees learned partly by doing relatively basic work. They produced the first version. Someone more experienced reviewed it and explained why making changes could make it better.
You learned what mattered and what didn’t. You developed a feel for language, people, risk, and context. It was your apprenticeship of work.
If AI increasingly performs the first draft, undertakes the initial research, and provides the basic analysis, what replaces that learning experience?
We will end up with a generation of employees capable of producing more sophisticated work much earlier in their careers, but with fewer opportunities to develop the underlying skills required to know whether that work is actually any good.
I am optimistically holding firm to the belief that the human skills of critical thinking, curiosity, judgement, empathy, and the confidence to challenge and understand nuance and context are not about to be replaced by AI. Because those things aren’t developed by reading about them, they are developed through experience, making mistakes, having difficult conversations, disagreement, and occasionally struggling with something that doesn’t have an immediate answer.
How can employers protect critical thinking as AI use grows?
Employees are already noticing the effect. A 2026 IBM Institute for Business Value study of 1,500 CHROs and 8,800 employees found that 60% of employees worry AI is eroding their skills. Critical thinking was the skill they cited most often as declining. The same study found that CHROs rank critical thinking and human judgement among the most important capabilities in their workforce.
So the challenge for employers isn’t about teaching people how to use AI, it’s about how we teach them to think without it.
And this doesn’t mean banning the use of AI. That would make about as much sense as banning calculators or using the satnav in your car. (I would be perpetually lost.)
The positive opportunities AI presents us are enormous. It can save us huge amounts of time. But efficiency is not the only measure of value.
Sometimes the process of getting to an answer is what makes us capable of making better decisions and being more effective.
So, my message is that in our rush to embrace the positives of AI, let’s make sure we continue to value and practise the things that develop our own judgement, creativity, curiosity, and ability to think—the “human bits”.
Frequently asked questions
Q: Does using AI reduce critical thinking?
A: It can when AI replaces the work that builds judgement. Critical thinking develops through experience, making mistakes, and struggling with problems that do not have an immediate answer. Using AI to accelerate your thinking is different from using it to avoid thinking.
Q: Why are early-career employees most affected by AI?
A: Early-career employees have not yet developed the judgement to know whether what AI tells them is right. They also have fewer chances to learn by producing first drafts that experienced colleagues review.
Q: What is the apprenticeship of work?
A: The apprenticeship of work is how junior employees learn by doing basic work that someone more experienced reviews and explains. It builds a feel for language, people, risk, and context.
Q: Should employers ban AI to protect critical thinking?
A: No. Banning AI would make as little sense as banning calculators. Employers should keep valuing and practising the tasks that develop judgement, creativity, curiosity, and the ability to think.
