Helping Your Team Embrace AI:

You’re unlikely to ever hear a colleague say “I have a fixed mindset about AI”. We just don’t see or describe ourselves that way. What you’ll hear instead are statements like these, each one rooted in our fears, past failures, ego and inhibitions.

"I'm not really a tech person."

An identity statement. It closes the door before the conversation has started.

"I'd rather just do it properly myself."

Quality standards borrowed as a reason not to learn.

"It's fine for the simple stuff, but not for what I do."

The last defensible position before somebody has to actually try it.

"If I need AI to do my job, what does that say about me?"

The real one. Almost never said out loud, and the one driving the other three.

Carol Dweck's contribution to all this came down to one small word: yet. "I don't know how to use it" becomes "I don't know how to use it yet". It sounds like a tiny distinction, but one that creates a powerful shift in perception. In a team, it is the difference between somebody turning their screen away and somebody asking a question about 'what could be'.

A study led by Elizabeth Canning with Mary Murphy, Jennifer Chatman, Carol Dweck and Laura Kray surveyed more than 500 employees at major companies and compared growth-mindset organisations with what they called cultures of genius. The differences were far from insignificant.

People in growth-mindset organisations collaborated more, took more intellectual risks, innovated more, and reported higher trust in and commitment to their employer. People in fixed-mindset organisations were more risk-averse, less collaborative, and more willing to cut corners and withhold information, because standing out mattered more than getting it right.

Now hold that next to AI. A fixed-mindset team hides its AI use, over-claims its own work, and won't admit what it doesn't understand. Which is exactly the behaviour KPMG and the University of Melbourne found across 48,000 people in 47 countries: 57% concealing their AI use and presenting the output as their own.

That is not a workforce with an integrity problem. That is a workforce that has read the culture correctly and responded sensibly. Which means the fix isn't a policy about honesty. It's a change in what feels safe to admit.

1. Surface the fear before you promote the tool:

You can't reframe something that hasn't been said out loud. Run one honest session with a single question: what do you think AI will take from us in this business? Write every answer on flip charts without defending any of them. Job security. Craft. Pride. Being the person people come to. The enjoyable bit of the work. Until those are visible, every training session you run to upskill on AI will be riddled with apprehension.

2. Swap the question:

Then ask the second one: what do you think you'd get back? Time, headspace, the two hours a week lost to formatting the report nobody reads. This is the pivot from uninterested to interested, and it has to come from them. If you supply the answer, it's a sales pitch. If they supply it, it's engagement.

3. Shrink the first step:

Start with one task people already find tedious and let AI take the first pass. Name the benefits, be honest about the risks, and make it clear the person stays in charge of the output. AI works on our terms. It isn’t a beast at the door. 

4. Learn out loud, failures included:

Fifteen minutes in a meeting you already hold. Three questions: What did you try? How did you audit the answer? Where did it let you down? Give the third question as much airtime as the other two, and go first yourself. A team that only shares the wins teaches everybody that struggling is a private matter, which is precisely the culture you are trying to leave behind. Curiosity is contagious, but only when it is visible.

5. Spend the time you get back on what makes a difference:

BCG found 42% of regular frontline users now save at least eight hours a week, and 66% get limited or no guidance on what to do with those hours. Over half don't redirect them into anything strategic at all. If your team cannot see where the time went, AI just feels like a treadmill that got faster. Decide together what the reclaimed time is for and protect it in the diary before it quietly evaporates.

Run steps one and two once. Run steps three, four and five on repeat. This is a habit you are building, not a project you are finishing.

One team, four very different reasons for hanging back. Insights Discovery gives you the language to coach each of them properly rather than delivering the same pep talk four times and wondering why only one person moved.

1. Colleagues with Fiery Red energy:

They resist anything that slows them down or risks making them look less than competent. Frame AI as a speed and control advantage rather than a learning curve, and give them a result to beat. They won't ask for help, so build the support into the task.

2. Colleagues with Sunshine Yellow energy:

They will dive in first and become your best early advocates, then drift once the novelty wears off. Give them a visible role that keeps them involved: let them run the first 15 minutes of a meeting and collect what everyone else is trying to do.

3. Colleagues with Earth Green energy:

They will worry about what this means for colleagues and for the human part of the work, and they will not say so in a big meeting. Ask them privately, take the concern seriously rather than talking them out of it, and show them where AI protects people's time instead of replacing them.

4. Colleagues with Cool Blue energy:

They will hold back until they trust the accuracy and know the rules. Don't push. Give them the policy, the checking standard, and a well-defined task where they can verify the output themselves. Once they are convinced they become the quality backbone of the whole thing.

Same message, four doors. If you already have colour energies mapped across your team, you know which door to knock on.

The framing matters far more than the equipping. A team that believes AI has come to take something will use it defensively, privately and badly. A team that believes it has come to bring something in return will use it openly, and openly is the only way anybody gets any good at it.

The statistics back the mood. BCG found more than two thirds of regular AI users report higher job satisfaction, and Slack found daily users are 81% more likely to report strong job satisfaction than people who never touch it. That is not the technology doing that. That is what happens when somebody stops feeling like they are behind.

Worth holding one caution alongside this. BCG also found 41% report increased mental strain, and 60% say the bar for what counts as good enough work has gone up. Curiosity without an honest conversation about workload just produces a faster treadmill and a more tired team. Step five isn't optional.

Your job in all of this is not to be the most AI-literate person in the room. It is to be the person who makes it safe not to be. That's team development far more than it is technology work, and it sits right alongside the coaching leadership style we wrote about earlier this year: ask more than you tell, and let people find their own reason to be interested.