Most organisations have already crossed the AI adoption line. Budgets are approved. Tools are rolled out. Teams have been told to get on board.

But something is not adding up.

A new study from the National Bureau of Economic Research (NBER) surveyed nearly 6,000 senior executives across the US, UK, Germany, and Australia. The headline finding: 69% of firms are actively using AI. That sounds like progress.

But look closer, and the real story is about the people at the top.

The AI Adoption Gap Starts With Leadership

The NBER data reveals three ironies that expose a deeper problem.

First, leaders are betting big on AI, but they barely use it themselves. The average executive spends about 1.5 hours a week on AI. That is roughly 18 minutes a day.

And 28% of senior executives do not use AI at all during a typical working week. These are CEOs, CFOs, and senior finance managers at firms that have already adopted the tools.

Second, the teams worried about AI replacing their jobs are the ones actually using it. The people closest to the tools are doing the work. The people making the decisions are watching from a distance. Within the C-suite, CEOs tend to use AI more frequently than CFOs and other senior executives, which means the credibility gap is not uniform. It concentrates in specific leadership roles.

Third, the “anxious” workforce is more optimistic about AI’s future than the executives pushing it. Employees predict AI will increase employment by 0.5% over the next three years. Executives predict AI will cut jobs by 0.7%, and by 1.2% at US firms.

That is not a small gap.

That is a fundamentally different story about the future.

From Adoption Gap to Credibility Gap

This goes beyond whether leaders are technically adopting AI. The real issue is credibility.

When leadership does not use the tools they mandate, it sends a clear signal to teams: “This is for you to figure out, not me.”

And teams notice.

The NBER study also found that more than 90% of firms report no material impact of AI on employment or productivity over the past three years. Firms are adopting the tools, but not yet seeing meaningful results. That is worth pausing on. If leaders are not deeply engaged with AI themselves, it is difficult for organisations to move past surface-level adoption and into genuine workflow integration.

Across Asia-Pacific, the pattern is familiar. The strategy deck says “AI-first.” The town hall says “We’re investing in AI.” But when leaders are asked, “What did you use AI for this week?” the room goes quiet.

Employees believe AI will create more opportunity. Leaders are planning for fewer people. That disconnect does not build trust. It erodes it.

Why Another Policy Will Not Fix This

The fix is not another AI policy or a bigger training budget.

The fix is leaders going first.

Here is what that looks like in practice:

• Use it visibly. Let your team see you experimenting, learning, even failing with AI. When leaders model behaviour, it normalises adoption for everyone else.
• Close the credibility gap. If you are asking your team to change how they work, change how you work first. The AI adoption gap shrinks when leaders participate, not just approve.
• Match your optimism to your action. Your team already believes AI can help them. The question is whether they believe you believe it too.

The NBER data is clear: adoption is not the bottleneck anymore. 69% of firms are in. The real question is whether leaders are willing to lead from the front, or keep leading from the slide deck.

Closing Thoughts

Your team is not waiting for permission. They are waiting for proof.

The AI adoption gap will not close with more announcements, bigger budgets, or better slide decks. It closes when leaders show up as users, not just sponsors.

The good news: this is one of the simplest leadership shifts to make. It does not require a new programme or a new tool. It requires 18 minutes a day and the willingness to be seen learning. And the trend is moving in the right direction. The NBER data shows that executive AI usage increased by roughly 50% between early 2025 and late 2025. The gap is closing, but slowly. The leaders who move first will set the pace for everyone else.

If your organisation is navigating this gap between executive intent and team-level adoption, Radiant Institute works with leaders and L&D teams to build practical, workflow-first AI enablement that starts where it matters most: at the top. Reach out to explore how we can help.