Leading a Team Through AI-Driven Change When You Are Uncertain

You do not need to have all the answers. You need to stop pretending you do.

That is the core of it. The delivery leaders I have seen handle AI-driven change well did not handle it by becoming AI experts overnight, or by projecting confidence they did not have. They handled it by being honest about what they knew, clear about what they were working to understand, and deliberate about not letting their own uncertainty become their team's problem to manage.

This is the version of the conversation that most leadership content skips. Everyone wants to talk about AI strategy and change management frameworks. Fewer people want to talk about what it actually feels like to stand in front of a team that is quietly anxious about what AI means for their roles, when you are quietly anxious too.

The situation is more common than leaders admit

McKinsey's January 2025 workplace report landed a finding that should give every delivery leader pause: the biggest barrier to AI adoption in organisations is not employee resistance. Employees, by and large, are ready. The biggest barrier is leadership. Leaders are not moving fast enough, not because they lack the tools, but because many are operating in the same fog their teams are, without saying so.

In parallel, 35 percent of employees in the same survey cited workforce displacement as a concern about AI. That number is not abstract. It represents people sitting in your programme, on your team, in your stand-ups, quietly calculating what all of this means for their next two years.

You are being asked to lead people through a change whose full shape you cannot yet see, using technologies that are genuinely evolving faster than any individual can track, toward outcomes that organisations are still figuring out how to define. The uncertainty is not a gap in your preparation. It is the nature of the situation.

What you do with that uncertainty is what matters.

What most leaders do instead

The default under pressure is performance. Look confident. Have an answer. Project direction. When a team member asks what AI means for the way the team works in twelve months, the temptation is to reach for something reassuring, something that sounds like a plan even when the plan does not yet exist.

This is where most leaders make the mistake that costs them later.

Research published in MIT Sloan Management Review found that leaders who openly express uncertainty and actively seek team input in resolving ambiguous challenges produce teams that are more satisfied, more committed, and less cynical. The counterintuitive truth is that performed certainty erodes trust faster than admitted uncertainty. Your team are experienced people. They know when they are being managed rather than levelled with.

A senior delivery leader I know found himself in exactly this position twelve months ago. His organisation had decided to introduce AI-assisted tooling into the development workflow. He had been involved in the decision, believed in the direction, and had no clear answer to what it meant for two of his most experienced engineers who did the kind of work the tooling was designed to support. He workshopped the messaging, tried to get ahead of the anxiety, and delivered something smooth that landed badly. The engineers did not ask more questions. They went quiet. He later understood that what they needed was not a polished answer. They needed to know he was taking the question seriously. They ended up leaving for another company.

The honest version

Leading through uncertainty is not about projecting false clarity. It is about being deliberate with the clarity you do have, and honest about the rest.

There are usually three things a delivery leader knows, even in the middle of genuine uncertainty about AI change.

What is actually changing, now. Not in twelve months, not eventually. Right now. What is the team being asked to do differently this quarter? What decisions have already been made at levels above yours? Communicate that clearly, without inflation and without minimising. People can handle hard information. What they struggle with is the sense that information is being withheld.

What is not changing. The anxiety about AI in your team is often not really about the technology. It is about role, identity, relevance, and belonging. Many of those things are not changing, or not changing in the way people fear. Say so directly. Not as reassurance, but as fact. "The judgement you bring to how we scope and prioritise this programme is not something a tool replaces" is more useful than "don't worry about AI."

What you are working to understand. This is the part most leaders skip. Naming what you do not yet know, and what you are doing to find out, is not a display of weakness. It is a display of honesty that your team will read as trust. "I do not yet have a clear view of what this means for how we plan the next phase. I am in conversations with the CTO and the product team, and I will share what I learn as soon as I have something concrete" is more credible than a holding answer dressed up as direction.

What this looks like in practice

Three things that make a material difference.

Separate your uncertainty from your team's anxiety. Your personal uncertainty about AI is yours to manage. Your team's anxiety is yours to address. These are related but not the same. Go and develop your own view: read, talk to people who are closer to this than you, test the tools yourself, have the uncomfortable conversations with your peers. Your team does not need you to have resolved your own uncertainty before they can feel stable. But they do need you to be actively working on it, not performing as if it does not exist.

Create the conversation rather than avoiding it. In a programme I ran where AI was being introduced into the quality assurance process, the anxiety did not surface in the team meetings. It surfaced in the one-on-ones, three weeks in, when two people independently described the same concern using almost identical words. The signal was obvious in retrospect: they had talked to each other before they talked to me. If the conversation is happening without you, it is already shaping the team dynamic. Surface it. Name it. Ask what people are thinking, not as a management technique but as a genuine question. PwC's 2025 Global Workforce survey found that without trust, employees are less likely to believe leaders' narratives about AI, and that trust is built through candour, not communication strategy.

Distinguish between uncertainty and instability. These are not the same, and your team needs you to hold the difference. You can be uncertain about what AI means for how roles evolve over the next two years and still provide a stable, clear direction for the next quarter. Uncertainty about the horizon is not the same as having no idea where you are going this week. Anchor your team in what you can control: the current sprint, the upcoming milestone, the decision that is on the table right now. Deliver on those. The trust that builds in the short term is what creates tolerance for genuine uncertainty in the longer term.

The question delivery leaders need to be asking is not "how do I reassure my team about AI." It is "how do I lead honestly through something I am also figuring out, without making my uncertainty their burden."

The answer is not a framework. It is a practice. Say what you know. Say what you do not know. Do the work of finding out. Show your team that you are taking their concerns seriously enough to think carefully about them, rather than managing them away.

That is what leadership in genuinely uncertain conditions actually looks like. It is less polished than what most leadership content suggests. And it is considerably more effective.

Meriem Debbabi is a Digital Transformation Leader based in Dubai. She writes about delivery excellence, high-performing teams, and the human side of change. Subscribe to Decoded, her monthly newsletter on AI and enterprise tech.

Sources used

McKinsey & Company, "Superagency in the Workplace," January 2025: leadership as biggest barrier to AI adoption; 35% of employees cite workforce displacement as a concern. https://www.mckinsey.com/capabilities/tech-and-ai/our-insights/superagency-in-the-workplace-empowering-people-to-unlock-ais-full-potential-at-work

MIT Sloan Management Review, "Embracing Uncertainty": leaders who express uncertainty and seek input produce more committed, less cynical teams. https://sloanreview.mit.edu/article/leadership-embracing-uncertainty/

PwC Global Workforce Hopes and Fears Survey 2025: trust as the precondition for employees believing leaders' AI narratives. https://www.pwc.com/gx/en/issues/workforce/hopes-and-fears.html

Bordia et al., Journal of Business and Psychology, 2004: management communication reduces strategic uncertainty; participation in decision-making reduces structural and job-related uncertainty. https://link.springer.com/article/10.1023/B:JOBU.0000028449.99127.f7