My staff are frightened about their jobs. How do I lead this change honestly?

I can feel it in the room. People smile in the AI workshop, lean in, ask the right questions, and then go quiet, because what they are really asking is whether I am training them to make themselves redundant. You catch it in the half-second pause after someone says "this is great," and you carry it home, because you are the one who booked the workshop.

So here is the answer, plainly. When your employees are afraid AI will take their jobs, the change-management move that works is not reassurance. It is candour paired with a visible reskilling path. Tell them the truth about what is changing, show them the route from where they are to where the work is going, and put names and dates against it. Comfort without a plan reads as a cover story. A plan, even an imperfect one, reads as respect.

Why does telling them "your job is safe" make it worse?

Because they have already decided you might not be telling the truth. In the 2025 Edelman Trust Barometer flash poll (Edelman is a global communications firm whose annual barometer measures public trust across dozens of countries) across five nations, two-thirds or more of respondents in developed countries said they believe business leaders will not be fully honest with employees about AI's impact on jobs; in the United States that figure reaches 70%. Sit with that. Before you open your mouth, the room has priced in a discount on your sincerity. So the blanket reassurance, "nobody is losing their job to this," lands as exactly the kind of thing a leader says when the opposite is true.

The fear is not abstract, either. Pew Research found 52% of US workers are worried about the future impact of AI in the workplace, and 32% think it will mean fewer job opportunities for them over time, against just 6% who expect more. A third (33%) feel overwhelmed. This is the weather your people now work in daily. The smile in the workshop is real; so is the calculation underneath it.

And the fear is climbing while leaders look away. Mercer's Global Talent Trends 2026 report, drawn from nearly 12,000 executives, HR leaders, investors and employees, found the share of employees who fear losing their job to AI has jumped to 40%, up from 28% two years earlier. The same report names the gap that matters most here: 62% of employees say leaders underestimate the emotional and psychological toll of AI, yet only 19% of HR leaders build that toll into their rollout plans. People are already using the tools, already worried about the tools, and largely carrying both in silence. The room is not confused. The room is being careful.

What does candour actually look like when I do not know who stays?

It looks like naming the shape of the change without pretending you can name every outcome. You do not have to promise certainty you cannot deliver. You have to be straight about the direction and specific about the support. The World Economic Forum's Future of Jobs Report 2025 gives you the language: employers expect 39% of workers' core skills to change by 2030. Not 39% of jobs deleted. Skills changing. That is a sentence you can say truthfully, because it is accurate and it is survivable.

Then you attach the path. In that same report, 85% of employers plan to prioritise upskilling, 50% plan to move people from declining roles into growing ones, and yes, 40% expect to reduce staff where skills stop being relevant. Candour means saying all three out loud, in that order, to the same room. The part most leaders skip is the hard arithmetic: of every 100 workers, 59 will need training by 2030, and 11 will need it but are unlikely to receive it. The fear in your workshop is the fear of being one of the eleven. Your job is to make the path to the fifty-nine real, named, and dated.

This is why the energy belongs on what becomes possible. PwC's 2025 global survey of nearly 50,000 workers found daily AI users were far more likely than infrequent users to report improved job security, 58% versus 36%, and higher salaries, 52% versus 32%. The people who get their hands on the tools feel safer, not less safe. That is the door you are opening: not "keep your job," but "become the person this work now needs." Point there.

Who do they actually believe, if not me?

Their peers. Edelman's 2025 poll found that "someone like me" is on average twice as trusted as a CEO or a government leader to tell the truth about AI. And in the UK and US, two in three of the people who distrust AI feel it is being forced on them. Read those two findings together and the strategy writes itself: your credibility now travels through colleagues, and your people want to feel like participants, not subjects of a rollout.

So set down the burden of being the single voice of reassurance, and start building a relay. Let your early movers, the ones already quietly using the tools, become the trusted "someone like me." Give them room to say what worked and what was clumsy. When the message comes from a peer who has used the thing and kept their job and got better at it, the discount on sincerity disappears.

What do I actually do on Monday?

Treat this as the change problem it is, not the technology problem it pretends to be. RAND, the US non-profit research institute, studied why artificial-intelligence projects fail and found more than 80% of them collapse, roughly twice the failure rate of conventional IT projects, and that the leading cause is not the model at all; it is leaders and teams misunderstanding or miscommunicating the problem the work is meant to solve (RAND, 2024). The thing keeping you up at night, the quiet room, the trust gap, sits squarely in that human layer. The bottleneck is no longer the technology. So lead the part the tools cannot lead on your behalf.

1. Name it first. Open your next session by saying the silent sentence out loud: "Some of you are wondering if I am training you to replace yourselves." Naming the fear is what proves you can be trusted with it.
2. Show the path, with dates. Commit to the reskilling route in public. WEF's number, 59 of every 100 workers needing training by 2030, becomes your promise that everyone on your team has a route to the growing roles, not the eleven left behind.
3. Hand the mic to peers. Let early movers tell their own story. "Someone like me" carries trust you cannot manufacture from the top.
4. Separate the two truths. Where roles are truly declining, say so early and offer real transition support. Candour about the 40% is what makes your optimism about the rest believable.

Think of the arc in three phases. Phase One, the Age of Effort: work hard, get a little more, linear growth. Phase Two, the Age of Scale: build once, sell to millions, exponential growth. Phase Three, the Age of Acceleration: output decoupled from human effort almost entirely, the phase AI now unlocks. Phase One was muscle. Phase Two was machine. Phase Three is mind, and this is a Phase Three problem: the work is to keep your people coherent (head and heart aligned, calm under pressure) and capable through a change they did not choose. Do that well and the smile in the workshop becomes the real thing.

About the author

Thomas Green

British technology futurist, AI keynote speaker and advisor. Thirty years across enterprise technology and AI strategy, helping leaders navigate the future of work. The futurist who died.