Are Your Best People Holding Back AI Adoption?

Leadership  Productivity
16 November, 2026

To accelerate AI adoption, change what it means to succeed

Think about the people we consider most successful in our organisations. They’ve built deep expertise, strong track records and considerable influence. We’ve rewarded them with greater responsibilities and relied on them to help build our businesses.

Now we’re asking these same people to embrace AI, which challenges the things that made them successful. Yet we continue to measure their performance and reward their contributions in ways that can discourage the changes we’re asking them to make.

In a cross-industry survey of 1,600 American knowledge workers and executives, 31% of employees admitted to actively working against their company’s AI projects.

We tend to assume that people resist AI because of skill gaps or fears about losing their jobs. But what if we’re overlooking another reason? Could our most valued and experienced people be slowing down AI adoption because it challenges the things that have helped them succeed?

This week, let’s explore why our best people struggle with AI adoption, how our management practices contribute to that resistance, and what we can do differently.

When enthusiasm fades

If we ask who has the most to lose when work is redesigned around AI, the answer isn’t always frontline staff or new hires. It could be the person with the longest track record, the biggest team, or the most exclusive knowledge base. It could even be the person leading the AI adoption initiative!

This is not to say that seniority automatically means resistance. A recent survey by SHRM found that 74% of older workers are keen to learn AI tools, while a BCG study found that 85% of leaders and 78% of managers use generative AI regularly.

Still, experience and expertise can play a role in sparking pushback to AI. Sometimes the people with the most to contribute to AI expansion also feel they have the most to lose. Consider the following implications:

  • Loss of identity and status. For many professionals, expertise is closely tied to a sense of self. People build their confidence and reputation on what they know – a knowledge often earned through years of experience. AI threatens to change this equation by making expertise more accessible. Researchers at Harvard Business School call it a “self-disruptive technology” because it can enhance results while making users feel less expert or visible in their jobs.

Think of someone who has spent 20 years becoming the person others consult. Now, anyone can get more information on their subject area from AI. What does this mean for the senior expert? The knowledge gap doesn’t disappear but it certainly becomes less visible.

  • Shrinking power and control. Management positions come with budgets, teams, reporting lines, and decision rights. With AI in play, these existing structures can no longer be taken for granted. If a team of 10 can complete much of its work with 5 people supported by AI, the manager’s role will change too. As more people rely on machines for answers, managerial influence is likely to wane.

The HBS researchers call this “span erosion” – managers keep their titles while losing a chunk of their traditional authority. This transition will be especially difficult in organisations that have historically treated team size as a proxy for managerial importance, since reducing headcount will automatically feel like a loss of power.

  • Lagging incentives. Companies continue to reward managers for predictable performance while simultaneously charging them to reinvent how work is done. Asking people to upskill, experiment with unproven tools, and overhaul their function while still hitting their quarterly numbers is a heavy load. We can’t expect employees to embrace a new operating model while evaluating performance using the old one.

According to Gartner, at least 50% of generative AI projects were abandoned after proof of concept by the end of 2025. The HBS researchers suggest that this failure could be partly driven by employees who see AI as a threat to their career prospects.

What leaders can do differently

Here are six suggestions for leaders to address AI resistance from experienced team members and managers:

1. Update what gets rewarded.

If we want people to embrace AI, our appraisals need to catch up. Let’s recognise managers who simplify processes, share expertise, and improve outcomes over a longer timeline – instead of focusing only on headcount and short-term targets. These changes should show up in performance reviews and promotion decisions, not just townhalls. Otherwise managers will keep protecting the structures and processes they have spent years building.

2. Make reinvention a career opportunity.

Help employees understand how they can continue their success through AI. “You can be more strategic” is too vague. People need to see a credible picture of what their contribution might look like. Could they move from managing routine analyses to advising on complex decisions? From personally preparing reports to providing human oversight for AI report-generation systems? Make the pathways tangible, with development plans and training opportunities for moving into new positions.

3. Lead by example.

Senior leaders can’t ask their teams to do something that they themselves refuse to do. Use AI to challenge your own way of working. Be open about it: share what worked, what didn’t, and what you’re still learning. As leaders, we need to demonstrate genuine AI proficiency rather than simply delegating adoption.

4. Make room to experiment.

Give people a safe place to test new ways of working, even at the cost of some disruption. PwC runs a dedicated AI “playground” and holds “prompting parties” for colleagues to swap strategies. Celebrate thoughtful attempts to redesign work, regardless of the result.

5. Give people a stake in the change.

A Harvard Business Review article highlights the importance of transparency and employee participation for successful AI adoption. Acknowledge the uncertainty around AI, then empower team members to join the process, influence decisions, and share accountability. This creates a much-needed sense of ownership. As the authors note:

In rapidly evolving workplaces, leaders who can satisfy employees’ need to feel competent, in control, and connected are able to foster adaptability, learning, and well-being on their teams.

6. Redesign the role before automating tasks.

Companies that rethink their workflows as a whole – instead of simply adding AI tools – report better efficiency, training outcomes, and employee engagement. Instead of leading with automation, begin by defining the new tasks people can take on once some of their routine responsibilities are delegated to AI. This gives team members continued purpose and motivation.

Getting people on board

Our most successful people bring experience, relationships, judgment and organisational knowledge that are essential to AI adoption. But we’re also asking them to reconsider some of the things that have defined their success, while continuing to reward those same things.

Which brings us back to the executive in the opening meeting. The next time a leader champions AI, take the conversation further. Ask what it means for their own function. What would they stop doing? What responsibilities would they give up? How would their role evolve? And what needs to change in the organisation to support them?

Those answers will tell us more than another presentation on AI adoption rates. They’ll also help us understand where the real resistance lies and what we need to do about it.

Are we giving our leaders enough reasons to embrace AI, or too many reasons to resist it?

X

Join the 8AM conversation

Loading