AI Adoption Mostly Comes Down to Whether You've Made It Safe to Try

There’s a pattern that keeps showing up in conversations with people leading teams right now. The company buys the AI tools and runs the training sessions and sends the all hands announcement. And then two weeks later the workflows are exactly the same and nobody can explain why.

It doesn’t matter what industry or how good the tools are. Something is clearly missing and it isn’t the technology.

The easy answer is that employees are afraid of change or don’t understand AI well enough yet. Most of the time that explanation is a cope. There’s something else going on.

Manager engagement globally fell from 31% to 22% between 2022 and 2025, according to Gallup. That’s a steep drop and it matters here because research consistently shows that the single strongest predictor of whether employees actually use AI at work isn’t the training program or the tool subscription. It’s whether their line manager actively champions it.

When a leader passes down a mandate they only half believe in, people read it immediately. They aren’t resisting the technology. They’re reading the room, and the room is telling them this probably won’t matter.

Sometimes the room is right. A Gallup survey found that only one in ten employees in organizations using AI strongly agree it has transformed how work gets done. The rest are practicing what’s started to be called ‘AI performance theatre’, going through the motions while the real workflows quietly continue.

A 2025 review by Welch on leadership engagement in AI receptive organizations found that leaders who proactively communicate AI benefits, model openness to innovation, and create psychologically safe environments significantly increase employee receptivity. This isn’t the soft stuff people tend to dismiss. It shows up in actual adoption behavior.

Patnaik and colleagues (2024), drawing on data from 250 companies, found that transformational leadership is a crucial driver of AI adoption and that a leader’s ability to recognize and respond to the feelings and needs of their team matters just as much as their technical vision. Both things. Together.

Whether your team uses AI mostly comes down to whether you, as the leader, have made it safe to try and fail with it.

Try it publicly before you mandate it. Share your own AI experiments in team meetings and include the ones that went sideways. Show the output that was genuinely useful and the one that was confidently, embarrassingly wrong (it helps if you laugh about it). Nothing reduces ‘psychological safety’ barriers faster than watching your manager look slightly bewildered by a bad prompt.

Separate experimentation from performance measurement. If people believe that using AI badly will count against them somehow, they won’t use it at all. Create space for genuinely low pressure practice before results are expected. This can be informal, a Slack channel, a monthly demo session, whatever fits your culture.

Ask about friction rather than adoption. Instead of tracking whether people are using the tools, ask what’s getting in the way. Ten minutes of direct conversation will tell you more than any usage dashboard. Usually it’s something small and completely fixable that nobody thought to mention out loud.

Model intellectual humility about AI openly. Leaders who say “I don’t fully understand this either and I’m figuring it out too” give their teams explicit permission to not already know everything. That permission is worth more than most onboarding programs.

The organizations making real progress with AI right now aren’t necessarily the ones with the best tools or the biggest budgets. They’re the ones where someone in a leadership role decided to get curious out loud, in public, and made it feel completely normal for everyone else to do the same. It turns out that’s just a lot of the job.

Are you seeing this in your own teams? I’m genuinely curious whether the gap between AI mandates and actual adoption is coming down to manager behavior or whether something else is driving it where you are.

I write about leadership and AI intersections regularly over on LinkedIn. Come continue the conversation there.

---

Sources

• Leadership Engagement and Organizational Culture in AI Receptive Organizations (Welch, 2025, Land Forces Academy Review)

• Exploring Determinants Influencing Artificial Intelligence Adoption, Reference to Diffusion of Innovation Theory (Patnaik et al., 2024, Technology in Society, 26 citations)