For the past couple of years, there’s been a lot of talk about AI adoption – How do we encourage employees to use it? How do we build confidence and capability? Where should we experiment? How do we help people understand what AI can and can’t do?
All important questions. But as AI becomes more deeply embedded in the workplace, another challenge is beginning to emerge. Because there’s a big difference between asking employees to use AI and using AI to make decisions about employees.
And that difference brings us to trust.
AI is becoming part of how work is managed
The UK Government is currently consulting on the use of workplace monitoring technologies, including some AI-enabled systems. The consultation covers everything from productivity scoring and automated decision-making to biometric surveillance, keystroke monitoring and GPS tracking.
It’s an interesting sign of where the workplace technology conversation is heading.
Until relatively recently, much of the discussion around AI at work was about the tools employees themselves might use. ChatGPT, Copilot and other generative AI platforms made AI visible. Employees could experiment with them, decide when they were useful and, to some extent at least, remain in control of how they interacted with them.
But AI is increasingly appearing somewhere else too. Behind the systems and processes that shape people’s working lives.
It can help organisations screen candidates, analyse employee sentiment, allocate work, identify patterns in performance, predict workforce trends and recommend decisions. In some workplaces, technology can also track activity, location, productivity or behaviour.
There can be perfectly legitimate reasons for using many of these technologies. They can help organisations make better decisions, identify problems earlier and remove significant amounts of manual work.
But from an employee experience perspective, there’s another question we need to ask – What does it feel like to be on the other side of them?
Trust changes when technology starts making decisions about us
Imagine being told that your organisation is introducing AI to help improve workforce planning. That might sound fairly uncontroversial.
But what if you don’t know what information the system is looking at? What conclusions it is drawing? Who sees those conclusions? Whether they influence decisions about your performance or opportunities? Or what happens if the system gets something wrong?
The technology hasn’t necessarily changed. But your relationship with it has.
This is why the next stage of AI adoption can’t simply be about encouraging employees to embrace new tools. Organisations also need to think carefully about the conditions in which employees will trust those tools to become part of how work is organised and decisions are made.
And trust isn’t created by telling people that a system is safe, accurate or efficient.
It’s built through what organisations do around it.
Transparency matters, but it isn’t enough
One obvious part of this is transparency.
If AI is being used in ways that affect employees, people should have a reasonable understanding of where it is being used, why it is being used and what role it plays in decisions that affect them.
That doesn’t mean every employee needs a technical explanation of how every algorithm works. It means they shouldn’t be left wondering whether a system is monitoring them, analysing them or influencing decisions about them without their knowledge.
But transparency alone isn’t the same as trust.
An organisation could be completely transparent about the fact that it monitors every keystroke an employee makes. Employees would certainly understand what was happening. That doesn’t mean they would feel comfortable with it.
The more interesting questions are about purpose and proportionality. Why are we collecting this information? What problem are we trying to solve? Is this the least intrusive way to solve it? What will we do with the insight? And, importantly, would we be comfortable explaining all of that openly to the people affected?
Those questions move us beyond AI governance and into employee experience.
Employee voice needs to enter the AI conversation
This is where human-centred design has an important role to play.
When organisations introduce new workplace technology, employees are often involved relatively late in the process. A system is selected, configured and prepared for launch, then the focus turns to communication, training and adoption.
At that point, we’re essentially asking: how do we get people to use this?
A human-centred approach asks a different set of questions much earlier.
What problem are we trying to solve? Who experiences that problem? What do they actually need? How might this technology change their experience of work? What concerns might they reasonably have? What unintended consequences could it create?
And crucially, what do the people affected think?
Employee voice shouldn’t simply be something organisations use to understand how people feel about AI after it has been introduced. It can be part of how decisions about AI are made in the first place.
That might mean involving employees in identifying potential use cases, testing new systems, exploring risks, setting principles for responsible use or deciding where human judgement should remain central.
The objective isn’t to give every employee a veto over every technological decision. It’s to recognise that the people experiencing work often see risks, opportunities and unintended consequences that aren’t visible from a technology or procurement perspective.
Trust is built before launch
There is a temptation to treat trust as a communication challenge.
If employees are nervous about AI, perhaps we need to explain it better. Share more success stories. Provide training. Reassure people that AI is there to support rather than replace them.
Those things may help. But if the underlying decisions don’t feel trustworthy, better communication won’t fix the problem.
If employees believe technology is being used primarily to monitor them, reduce their autonomy or make opaque decisions about their future, an internal communications campaign about the benefits of AI is unlikely to change much.
Trust is built much earlier, through the choices organisations make about where AI should be used, where it shouldn’t, what data is appropriate to collect, where human judgement remains important and how much influence employees have over decisions that affect their experience.
The organisations that build trust may ultimately get more from AI
There is an interesting irony here.
Organisations understandably want employees to experiment with AI, share ideas and identify opportunities for it to improve work. But people are far more likely to engage openly with new technology when they understand how their organisation intends to use it and believe that those decisions will be made responsibly.
If employees worry that AI will be used to monitor them, assess them or make decisions they don’t understand, they may become more cautious rather than more curious.
That matters because some of the most valuable AI opportunities are likely to emerge from the people closest to the work. They know where time is being wasted, where information is difficult to find, which processes are frustrating and which parts of their jobs genuinely benefit from human judgement.
Creating the conditions for them to contribute those insights requires more than access to technology.
It requires trust.
As AI becomes a more ordinary part of working life, perhaps the organisations that make the most progress won’t simply be those with the best tools or the highest adoption rates. They’ll be the ones that involve people in shaping how those tools become part of work.
At People Lab, we help organisations use employee insight and human-centred design to make better decisions about work and employee experience. As technology continues to change what is possible, those principles become more important, not less. Because the question isn’t simply whether people are ready for AI. It’s whether we’re creating workplaces in which they can trust how it is being used.
