Critical Thinking Skills: How Teams Reduce Cognitive Bias at Work

Cognitive bias rarely feels like bias when it’s happening. It usually feels like common sense, speed or confidence. In this article, we explore how critical thinking skills help teams spot assumptions, challenge weak evidence and sense-check AI outputs before those shortcuts turn into workplace decisions.

How Do Critical Thinking Skills Reduce Cognitive Bias?

Critical thinking skills reduce cognitive bias by helping people pause before accepting their first judgement. They give teams practical ways to separate facts from assumptions, look for evidence that challenges their view and sense-check information before acting. They don’t remove bias completely, but they reduce the chance that important workplace decisions are driven by instinct, pressure, familiarity or overconfidence.

Sadly, we’re humans, not beautifully calibrated decision-making machines.
We like to think we’re rational at work. That decisions are based on data, experience and good judgement. And sometimes they are.

But more often than we’d like to admit, what’s really driving those decisions is a set of invisible shortcuts our brains are taking without us noticing. That’s cognitive bias.

It shows up in all the places you’d expect: hiring, performance conversations, strategy meetings, client decisions and project planning. But it also shows up in the smaller, everyday moments. The quick judgement on a piece of work. The decision to trust one data point over another. The assumption that the most confident person in the room must be right.

And now we’ve added AI into the mix.

Which means we’re not just dealing with our own bias anymore. We’re also dealing with how we interpret and trust AI-generated information. If anything, that’s made critical thinking skills more important, not less.

Because the real risk isn’t that people don’t have access to information.

It’s that they accept it too quickly.

In our workshops, the problem we see most often is not that people lack intelligence or care. It’s that they start challenging the decision after it has already gathered momentum. Critical thinking works best earlier, before assumptions harden into plans.

Quick Answer: How Critical Thinking Reduces Cognitive Bias

Critical thinking reduces cognitive bias by slowing down automatic judgement. It helps teams check evidence, identify assumptions, consider alternative explanations and challenge AI outputs before using them in decisions.

That pause matters.

Without it, people often make decisions based on what feels familiar, recent, easy, confident or convenient. With it, teams are more likely to spot weak evidence, challenge assumptions and make decisions that hold up under scrutiny.

What Is Cognitive Bias at Work?

Cognitive bias is a repeated pattern of thinking that can distort judgement, especially when people are under pressure, working quickly or relying on incomplete information.

It usually kicks in when we’re trying to make sense of something quickly. We lean on past experience, go with what feels familiar, fill in gaps, or give extra weight to information that confirms what we already believe.

That’s not a character flaw. It’s how the brain copes with complexity.

The problem is what it looks like in practice.

In a workplace, cognitive bias might mean:

• backing an idea because it’s similar to something that worked before
• giving more weight to the most recent piece of feedback
• trusting a confident opinion over a well-evidenced one
• hiring someone because they “feel like a good fit”
• accepting an AI-generated answer because it sounds polished and credible

None of those decisions feel biased in the moment. They feel efficient. Sensible, even.

That’s exactly why they’re so hard to catch.

What Are Critical Thinking Skills in the Workplace?

Critical thinking skills are the skills we use to question, analyse and evaluate information before making a decision.

That might mean checking where a piece of information has come from, separating fact from opinion, looking for missing context, or deliberately asking, “What else could be true here?”

In practice, critical thinking is less about being clever and more about being deliberate.

It’s the difference between: “That sounds right.” and “What makes that right?”

It’s also the difference between: “The AI gave me a useful answer.” and “The AI gave me a useful answer, but what might it have missed?”

That second question is where the value sits.

In our critical thinking workshops, we often see people realise that the issue isn’t a lack of intelligence or effort. It’s that they’re making decisions quickly, under pressure, using whatever information is easiest to reach.

And that’s when shortcuts sneak in.

Critical Thinking, Cognitive Bias and Heuristics: The Difference

These terms often get bundled together, so it’s worth separating them.

Heuristics aren’t automatically bad. We need shortcuts. Without them, we’d be paralysed trying to decide whether to open an email, make a coffee or put socks on. Although, to be fair, some mornings even that feels ambitious.

The problem is when those shortcuts quietly become the basis for important decisions.

That’s where critical thinking skills matter.

Why Critical Thinking Matters More in AI-Supported Work

One of the biggest lightbulb moments we see in our critical thinking workshops comes when we talk about how quickly people are making decisions.

There’s a statistic that gets thrown around saying the average person makes around 35,000 decisions a day. Whether that number is accurate is ironic and slightly beside the point. In fact, that’s part of the learning. It’s very easy to accept something as true just because it sounds credible, appears in a search result, or gets repeated often enough.

What we do know is this: we make a huge number of decisions every day, and most of them happen fast.

In Daniel Kahneman’s brilliant work on system 1 and system 2 thinking, he shares how hard, but essential it can be to switch out of automatic thinking.

Modern work gives people more information, more noise, more urgency and more pressure to respond quickly. That makes critical thinking harder to practise at exactly the moment we need it most.

At VTT, we’ve delivered 33 Critical Thinking and Critical Thinking for AI workshops since 2023 across sectors, industries and countries. Over that time, we’ve seen demand for critical thinking grow sharply, particularly as organisations look for practical ways to help people use AI confidently without accepting its outputs too quickly.

That demand is still growing. Bookings for Critical Thinking with AI in 2026 are already up 75% compared with 2025.

That shift makes sense. Critical thinking has always mattered in workplace decisions, but AI has made the gap more visible. People are not just weighing up human opinions, reports and data anymore. They’re also interpreting AI-generated summaries, recommendations and research.

Why More Information Does Not Mean Better Decisions

Another moment that tends to land heavily in workshops is when we look at the sheer volume of information being created.

The numbers are almost hard to get your head around. In 2010, the world generated around 2 zettabytes of data. By 2025, that figure was estimated at around 181 zettabytes.

That curve isn’t just growing. It’s accelerating with the estimate for the end of the decade to be closer to 550 zettabytes (Statistica).

But the bit that really makes people pause is this: more and more information is being generated, summarised or reshaped by AI.

So we have more information than ever, created faster than ever, increasingly by systems that sound confident and convincing.
And yet we’re still relying on the same mental shortcuts to process it.

That’s where the risk sits.

Not in the volume of information itself, but in how quickly we accept it.

AI Has Not Made Us Worse Thinkers. It Has Raised the Stakes.

There’s a growing narrative that AI is making people less intelligent, or that our critical thinking skills are in decline because of it.

We’re not convinced that’s the whole story.

It’s not that people suddenly stopped thinking critically the minute ChatGPT appeared. The bigger issue is that the need to think critically has increased, while the environment we’re working in makes it easier not to.

We’ve always had a tendency to trust information that feels authoritative.

Years ago, that might have sounded like: “It must be true. I read it in the newspaper.”

Now it sounds more like: “It must be right. The AI said it.”

Different source. Same human behaviour.

The challenge isn’t new. The scale and speed of it is.

Because AI outputs are often well-written, structured and confident, they can feel more reliable than they actually are. That doesn’t mean AI is the problem. AI can be hugely useful. But it does mean people need the skills to challenge it.

In our work with teams, one of the most common AI mistakes we see is not people missing obvious nonsense. They usually spot that. The bigger issue is that they miss what’s absent. The output sounds coherent, so they focus on whether it is useful rather than whether it is complete.

That is where critical thinking becomes essential so we are not over reliant on AI research (Microsoft).

A Practical Example: When an AI Summary Looks Useful but Misses the Quiet Signal

A common scenario we use in AI sense-checking work is the analysis of employee feedback, survey comments or open-text responses.

The AI output can look incredibly helpful. It takes messy qualitative feedback and turns it into clean themes, tidy bullet points and suggested actions. Lovely. Time saved. Everyone breathes out.

But when teams review that output using critical thinking, they often realise something important: the summary may be tidy without being complete.

For example, the AI might surface the most repeated themes, but miss less frequent comments that carry emotional weight. It might flatten differences between teams, locations or roles. It might summarise “what was said most often” but not “what might matter most”.

That’s where LEAP becomes useful.

This is the point. Critical thinking does not mean rejecting AI. It means using AI with better judgement.
A summary can be accurate and still incomplete.

That sentence tends to make people pause.

System 1, System 2 and Why Shortcuts Are So Hard to Interrupt

One useful way to explain critical thinking is through Daniel Kahneman’s idea of System 1 and System 2 thinking.

System 1 is fast, automatic and intuitive. It helps us make quick judgements without using much conscious effort.

System 2 is slower, more deliberate and more effortful. It’s the part we need when we’re analysing information, checking assumptions or working through complexity.

The problem is that modern work pushes us towards System 1 almost constantly.

Quick reply needed. Inbox full. Meeting starting. AI summary available. Decision required. Someone senior has already given their view.

So we default to speed.

Critical thinking is the conscious interruption of that default. It’s what helps people say, “Hang on, what are we assuming here?” before the group runs off in completely the wrong direction with a spreadsheet and misplaced confidence.

5 Cognitive Biases That Affect Workplace Decisions

Most people don’t set out to make biased decisions.

But bias doesn’t feel like bias when it’s happening. It feels like common sense.

Here are five biases that commonly show up in workplace decisions.

1. Confirmation Bias

Confirmation bias is the tendency to look for information that supports what you already believe, and give less attention to anything that challenges it.

A manager might already have a view that someone in their team is underperforming. When they look at the evidence, they focus on missed deadlines and overlook the projects that went well.

A project team might believe their proposed solution is the right one, so they pay more attention to customer feedback that supports it and quietly dismiss anything that complicates the picture.

With AI, confirmation bias can become even more subtle. If someone prompts an AI tool in a way that leads towards their preferred answer, the output may reinforce what they already thought. It feels like validation, but it might just be a beautifully formatted echo.

Critical Thinking Question to Ask
What evidence would change our mind?

2. Availability Bias

Availability bias happens when recent or memorable information has more influence than it should.

For example, a team might overreact to one recent customer complaint because it’s vivid and fresh, even if wider feedback tells a different story.

Or a senior leader might give too much weight to a recent problem because it was stressful, visible or escalated quickly.
We see this in decision-making meetings all the time. The most recent example gets the most airtime, regardless of whether it’s actually representative.

Critical thinking question to ask:
Are we reacting to what is most recent, or what is most relevant?

3. Framing Effect

The framing effect is when the way information is presented changes how people interpret it.
A project described as having a “70% success rate” will feel very different from one described as having a “30% failure rate”. Same data. Different emotional response.

In performance conversations, framing can influence how fairly someone is judged. “They struggled with this project” lands differently from “They delivered well in a difficult context, but needed more support in one area.”

With AI, framing matters because the way we prompt a tool shapes the answer we get back. Ask a narrow question and you’ll often get a narrow answer. Ask a leading question and you may get a very helpful-looking response that quietly follows your lead.

Critical thinking question to ask:
How would this look if it were framed differently?

4. Automation Bias

Automation bias is the tendency to over-rely on automated systems or recommendations, even when they may be wrong or incomplete.

In everyday work, automation bias might look like:

• accepting an AI-generated summary without checking the source
• assuming a recommendation is objective because a system produced it
• copying an AI-generated answer into a document without testing whether it’s accurate
• trusting a meeting summary even though you know the meeting itself was messy, nuanced or politically delicate

The output feels reliable because it’s confident, tidy and quick. That’s exactly what makes it risky.

Critical thinking question to ask:
What would we check if a human had given us this answer?

5. Familiarity Bias

Familiarity bias is the tendency to favour what we recognise.

At work, that might mean sticking with the same supplier, same process, same type of candidate or same strategic approach because it has worked before.

Sometimes that’s sensible. Experience matters.

But familiarity can also become a lazy shortcut. It can stop teams from noticing that the context has changed, or that a better option is available.

This is particularly relevant when teams are under pressure. When time is tight, people often reach for the familiar, because it feels lower risk.

Critical thinking question to ask:
Are we choosing this because it’s best, or because it’s familiar?

What This Means in Practice

A team does not need to stop using instinct, experience or AI.

That would be ridiculous. Also impossible.

The aim is to know when those inputs need checking.

Critical thinking gives people a practical pause point before a decision becomes action. It helps teams notice when they are being pulled by confidence, familiarity, urgency or a polished AI output, and gives them a way to challenge the thinking before it becomes the plan.

The Biggest Mistake: Teaching Bias as Awareness Only

Here’s where a lot of workplace training goes wrong.

It teaches people to name biases, but not to interrupt them.

Awareness matters, of course. People need to understand what confirmation bias, availability bias and automation bias are. But knowing the name of a bias doesn’t magically stop it influencing a decision.

That would be lovely. It would also make training design much easier, and we’d all have more time for lunch.
The real behaviour change comes from practice.

Teams need simple, repeatable habits they can use in live decisions. Not just in a workshop. Not just in an abstract exercise. But in meetings, reviews, project planning, AI use, client work and everyday conversations.

In our workshops, this is often the moment that shifts the room. People realise the issue isn’t that they never challenge decisions. It’s that they often challenge them too late, once the thinking has already narrowed and the preferred answer has started to harden.

That’s why critical thinking at work need to be practical.

What Most Teams Get Wrong About Critical Thinking

Most teams don’t struggle with critical thinking because people aren’t clever enough.

They struggle because critical thinking gets applied too late.

By the time someone asks a good question, the decision has often already gathered momentum. The preferred idea has been named. The senior voice in the room has spoken. The AI summary has made the issue look tidy. At that point, challenge can feel awkward, slow or unhelpful.

That’s why critical thinking needs to happen earlier, before a decision starts to harden.

In assumption-spotting exercises, teams often start with a confident recommendation and then realise within minutes that several of their “facts” are actually assumptions. They might be assuming what a client wants, why a colleague behaved in a certain way, or what a piece of data really proves.

That moment matters because it makes bias visible without making the conversation personal.
And that is the trick.

If critical thinking feels like criticism, people get defensive. If it feels like a shared way to improve the decision, people are much more likely to use it.

Use LEAP to Challenge Biased Thinking

At VTT, we use a simple model called LEAP to help teams challenge decisions before they harden into action.

LEAP stands for:

• Logic
• Evidence
• Attribution
• Perspective

It is the same framework behind our Spot-the-Flaw tool, which helps people challenge AI-generated answers rather than accepting them at face value.

LEAP works because it gives people four practical ways to challenge a decision, recommendation or AI output without turning the conversation into an argument.

One reason LEAP works well in our workshops is that it gives people something practical to use. In participant feedback, the value of “different models that can be applied to my work” came through clearly.

That’s the point. Critical thinking doesn’t become useful because people can define confirmation bias. It becomes useful when they have a simple way to challenge logic, evidence, attribution and perspective in the moment.

The model is deliberately simple. If a tool is too complicated, people won’t use it when they’re busy.

And busy is exactly when they need it.

Weak Decision Habits vs Stronger Critical Thinking Habits

This is not about slowing everything down.

It’s about spotting which decisions deserve more scrutiny.

Nobody needs a full critical thinking process for choosing biscuits for a meeting. Unless you’re buying Custard Creams, in which case, questions should be asked!!

Common Mistakes When Trying to Reduce Cognitive Bias

If you want teams to reduce cognitive bias, there are a few traps to avoid.

1. Teaching Bias as Theory Only

People need to understand bias, but theory alone rarely changes behaviour. The learning has to move quickly into real decisions, realistic scenarios and practical questions.

2. Assuming Awareness Equals Action

Someone can understand confirmation bias perfectly and still fall into it five minutes later. The point is not just to recognise the bias. It is to build habits that interrupt it.

3. Asking for Challenge Too Late

If challenge only happens after the preferred decision has already been shaped, it can feel like resistance. Critical thinking works best when it is built into the decision process early.

4. Treating AI Outputs as Neutral

AI outputs can look balanced and objective, but they are still shaped by prompts, data, sources and missing context. They need scrutiny.

5. Confusing Confidence with Evidence

This is a big one. A confident person, report or AI output can feel persuasive, but confidence is not the same as accuracy.

6. Assuming Experienced People Are Immune to Bias

Experience is useful. It can also become part of the problem if it turns into overconfidence or familiarity bias.

What We’ve Learned From Running Critical Thinking Workshops

Across our Critical Thinking and Critical Thinking for AI workshops, the most useful insights are rarely the most complicated ones.

They’re the simple shifts that change how people approach decisions.

1. People Usually Spot Bad Logic Faster Than Missing Context

When something is obviously wrong, people often catch it. A strange claim, a clumsy conclusion or a nonsense AI output will usually raise eyebrows.

What is harder to spot is what’s missing.

That’s why AI-generated content can be so tricky. It can sound coherent and still leave out important evidence, minority views, nuance or context.

2. Teams Often Challenge Decisions Too Late

This is the big one.

Once a decision has gathered momentum, challenge becomes harder. The preferred answer has social weight behind it. People don’t want to slow things down, irritate the room or look awkward.

Critical thinking works best before that point.

Before the assumption becomes the plan.

3. The Best Critical Thinking Questions Feel Practical, Not Academic

Nobody wants to be asked a question that sounds like it came from a philosophy exam.

Useful questions are simple:

• What are we assuming?
• What evidence supports this?
• What might we be missing?
• Whose perspective is absent?
• What would change our mind?

That’s why models like LEAP help. They give people a shared language for better questioning without making it feel heavy.

4. AI Confidence Makes Absence Harder to Notice

AI outputs can be fluent, structured and persuasive. That makes it easy to focus on what is there, rather than what isn’t.

That’s a different kind of critical thinking challenge.

The question is not only: Is this wrong?

It is also: What might be missing?

Critical Thinking Exercises for Teams

If you want people to build critical thinking skills, don’t just explain bias and hope for the best.

Give them something to practise.

Here are a few exercises we use, adapt or build into our work with teams.

Exercise 1: Assumption Spotting

Give the team a workplace scenario, decision or AI-generated recommendation.

Ask them to identify:

• what is known
• what is assumed
• what is missing
• what needs checking

This works particularly well because teams often realise how quickly they fill in gaps without noticing.

For example, a team might begin with the statement, “The client is unhappy because the project is behind schedule.” Within a few minutes, they may realise they know the client has asked for an update, but they are assuming the emotional state, the cause and the level of concern.

That’s a very different starting point.

Exercise 2: Red Team the Recommendation

Ask one group to build the strongest possible case for a decision.

Ask another group to challenge it.

The aim isn’t to “win” the argument. It’s to improve the decision by exposing weak evidence, hidden assumptions or missing perspectives.

This is especially useful for project decisions, strategy discussions and AI-supported recommendations.
When teams do this well, they often discover that the best challenge is not, “This is wrong.” It is, “What would need to be true for this to be right?”

Much less combative. Much more useful.

Exercise 3: Reframe the Issue

Take the same data and frame it in two different ways.

For example:

• “70% of people completed the programme”
• “30% of people did not complete the programme”

Then ask:

• What changes in how we interpret this?
• What emotional reaction does each version create?
• What extra information do we need before making a judgement?

This helps teams see how framing can shape decisions before the actual thinking has even started.

Exercise 4: AI Sense-Check Using LEAP

Give teams an AI-generated answer, summary or recommendation.

Ask them to review it through LEAP:

• Logic: Does the reasoning make sense?
• Evidence: What evidence is included, and what is missing?
• Attribution: Where has the information come from?
• Perspective: What viewpoint or context might be absent?

This keeps the conversation balanced. The point isn’t “AI is bad”. It’s “AI is useful, but we still need to think.”

That distinction matters.

How Managers Can Build Critical Thinking Into Team Habits

Critical thinking shouldn’t sit in a workshop and then quietly evaporate the minute people get back to their inbox.

Managers play a big role in making it part of the team’s normal rhythm.

That doesn’t mean turning every meeting into a philosophical debate. It means building small prompts into the way decisions are already made.

For example:

• In project meetings: “What are we assuming?”
• In performance conversations: “What evidence are we using?”
• In hiring decisions: “What does ‘good fit’ actually mean here?”
• In AI use: “What would we check before relying on this?”
• In team reviews: “What did we miss because we moved too quickly?”

The aim is to make critical thinking normal, not special.

Because if people only use it when someone books a training session, it won’t change much.

What Participant Feedback Tells Us

Since 2023, VTT has delivered 33 Critical Thinking and Critical Thinking for AI workshops across sectors, industries and countries. Demand has continued to grow, with 2026 bookings for Critical Thinking with AI already up 75% compared with 2025.

Participant feedback from these workshops has also been strong. Among those who gave feedback, 100% said they would apply the learning in their role, 94% said they expected to perform better as a result, and the average recommendation score was 9.1 out of 10.

The most useful evidence, though, is in what people say they will do differently.

Comments such as “I will question decisions more” and “Will adopt more reflection in decisions I make” show the shift we’re aiming for: not just understanding cognitive bias, but changing how people approach decisions.

Other comments pointed to the practical value of the training, including “different models that can be applied to my work” and “increased awareness to my thinking process”.

That matters because critical thinking does not become useful when people can define a bias. It becomes useful when they have simple ways to pause, question and test their thinking before assumptions harden into plans.

What This Means for L&D and HR Teams

For L&D and HR teams, the opportunity is bigger than simply helping people “think better”.

Critical thinking skills support better judgement across a whole range of workplace challenges: leadership, feedback, inclusion, AI adoption, decision-making, problem-solving and change.

They also help people build confidence in situations where there isn’t one obvious right answer.
That’s important, because modern work is full of those situations.

The value of critical thinking isn’t that it gives people a perfect process for every decision. It gives them a way to pause, question and improve the quality of their thinking when it matters.

That’s especially useful in organisations where people are moving quickly, using AI more often, and dealing with more information than they can realistically process.

In other words, most organisations.

Final Thoughts

Critical thinking skills are no longer a nice-to-have.

They’re becoming essential for teams that need to make better decisions in fast-moving, information-heavy and AI-supported workplaces.

If your teams are using AI more often, making complex decisions or struggling to challenge assumptions early enough, our Critical Thinking and Critical Thinking for AI workshops can help.

We use practical tools such as LEAP and Spot-the-Flaw to help people question information, sense-check AI outputs and make better decisions before weak assumptions become the plan.

Explore the workshops or get in touch to discuss what would be most useful for your team.

FAQs About Critical Thinking Skills and Cognitive Bias