AI is dividing your workforce. Champions are pushing ahead, sceptics are pulling back, and 31% of employees in a recent study admitted to working against their organisation’s AI strategy.
Closing that divide relies on one thing: trust.
Trust determines whether people lean into change or quietly opt out. Without it, even a well-funded AI strategy can remain exactly that: a strategy, understood in the boardroom but disconnected from what people do every day.
A helpful way to think about trust in an AI strategy is through the Head, Heart, and Hands framework. People need to trust the direction and understand what needs to change (Head). They need to trust the people leading it and have a reason to care (Heart). And they need to trust their own ability to take part, with practical guidance on what they need to do differently (Hands).
When one of these elements is weak, trust suffers. All three need to work together if an AI strategy is going to make it beyond the slide deck and into your people’s day-to-day work.
Head: Define what good looks like
Most organisations are comfortable with the Head because it produces visible things: strategy decks, town halls, policies, roadmaps, and messages from senior leaders.
All of that has a place. People need to understand where the organisation is heading and why. But clarity at an organisational level is very different from clarity for the person sitting at their desk wondering what any of it means for their job.
That is where an AI fluency map becomes useful. Rather than asking everyone to “use more AI”, define what good looks like role by role. Start with the work people do today, identify where AI could change it, build practice around those moments, and be clear about what would help the new behaviour stick.
For one person, that might mean knowing when AI can speed up research and when its answers need checking. For another, it could mean redesigning a regular task that takes hours each week.
Defining these specific ways of working also gives you something concrete to measure. Instead of focusing on whether people are using AI, based on licences, logins, or prompts, see if the changes you want are showing up in the work. You do not need a perfect data set to begin. In many cases, managers, team leads, and the people closest to the work can tell you whether a new behaviour is taking hold before a formal reporting system can.
By connecting AI fluency to observable changes in working practices, you give employees a clearer sense of what good looks like and leaders and managers a more credible way to see whether the strategy is taking hold. That builds trust on both sides: people know what is expected of them, while the organisation can see whether change is happening.
Hands: Create confidence to act
Once people know the direction, they need a safe way to move towards it. This is where governance matters. If people don’t know how to act safely, or fear the consequences of getting it wrong, they are unlikely to act at all. But governance has something of an image problem. It suggests policies, committees, and somebody somewhere saying “no”.
Done well, governance does the opposite. It gives people clear boundaries and, with them, permission to experiment. If I know which tools I can use, what information I can put into them, and when I should ask for help, I can get on with trying things. Good governance is therefore permission with shape.
People need guardrails that reflect the job they are doing and the work in front of them, rather than generic rules that leave them to work out what is safe on their own.
When people know where the boundaries are, they can act with more confidence. That builds trust because employees know what is safe to try, where the limits are, and where to turn when they are unsure.
Heart: Lead with trust
A strategy can succeed or fail based on whether people feel safe admitting that they don’t know something and whether they see leaders learning alongside them.
There is a big difference between hearing a senior leader talk about the importance of AI and seeing them use it in a meeting, question an answer it produced or admit when something did not work. Leaders who model this in the open give other people permission to do the same.
Then there are the harder conversations. People know AI could change jobs, tasks, and expectations. Avoiding the subject leaves a gap, and people often fill gaps with worst-case scenarios.
Leaders and managers don’t need to (and cannot) predict exactly what work will look like in five years. But they do need to be honest about what they know, what they don’t, and what they are doing about it. That honesty is part of what builds trust and makes the rest of the message believable.
Starting where trust is weakest
The Head, Heart, and Hands model can help organisations identify where trust is breaking down in their AI strategy. The goal is not to fix everything at once, but to find where the balance is off and start there.
If the problem is clarity on what good looks like in their role, work on Head. If people know what they should do but struggle to do it, look at Hands. If people understand the change but are still holding back, look at Heart.
When these three elements work together, employees have more reason to trust the strategy. When one is weak, that trust starts to erode.
To explore these ideas in more detail, watch our recent webinar, Head, Heart, Hands (and Gut): How to Rebuild Trust in Your AI Strategy. In the session, Callum explores more practical ways to build trust into AI adoption and help people change how they work.