I keep seeing the same pattern. A CEO comes back from a conference fired up about AI. They announce an AI strategy. They fund a pilot. They champion it publicly. Six months later, adoption sits at 15%. The CEO is frustrated. The team is confused. And the middle managers are caught between a directive from above and resistance from below. Gerson and I have been unpacking this disconnect, and it is more systematic than most leaders realise.
What You Need to Know
- CEO enthusiasm for AI does not translate into team adoption without deliberate mechanisms for bridging the experience gap
- The disconnect is not caused by poor communication. It is caused by a difference in exposure, context, and risk perception between executives and frontline teams
- Middle managers are the critical link, and they are typically the least supported stakeholder in AI adoption
- The fix is structural, not motivational. You cannot inspire people into adopting AI. You need to remove the barriers they face
15-20%
typical adoption rate 6 months into CEO-championed AI initiatives that lack structured middle management enablement
Source: RIVER advisory assessments, 2024-2025
The Experience Gap
What the CEO Sees
The CEO has been exposed to AI in a specific way: vendor demonstrations, conference keynotes, peer conversations, board presentations, consulting reports. In this exposure, AI is impressive, transformative, and clearly beneficial. The demos work. The case studies show ROI. The competitive landscape demands action.
This exposure is real but curated. It shows AI at its best, in controlled conditions, presented by people whose job is to make it look good.
What the Team Sees
The team's exposure to AI is different: a new tool they didn't ask for, in a workflow that was already under pressure, with training that was too brief, solving a problem they're not sure they have. The tool sometimes gives wrong answers. Nobody has explained what happens to their role. The metrics they're measured on haven't changed to account for the learning curve.
The CEO cannot easily imagine what it's like to use an AI tool when you're already behind on targets and suspect it might replace part of your role. The team cannot easily imagine the competitive pressure that makes AI feel urgent. The gap between these perspectives is where adoption stalls.
Dr Gerson Tuazon
AI Strategy & Health Innovation
What Middle Managers Experience
Middle managers get the worst of both worlds. They receive the directive to drive AI adoption from above. They receive the resistance and confusion from below. They're expected to champion something they may not fully understand, to a team that may not want it, with metrics that haven't been adjusted.
Most AI adoption strategies ignore middle managers entirely, treating them as a transmission mechanism rather than a stakeholder group with their own concerns, constraints, and needs.
Why Communication Doesn't Fix It
The default response to low adoption is more communication. Town halls, newsletters, success stories, executive videos. This rarely works because the problem is not information. The team knows the CEO wants them to use AI. They know it is strategically important. They know the future involves AI.
The problem is that knowing all of this doesn't address their actual barriers: insufficient training, unclear role implications, no psychological safety for admitting confusion, metrics that penalise the learning curve, and tools that don't fit their workflow.
What Actually Works
Enable Middle Managers First
Before launching AI to the workforce, equip middle managers with:
- Hands-on AI experience (not presentations, actual use)
- Honest answers about role implications
- Adjusted metrics that account for the adoption learning curve
- Authority to adapt the AI rollout to their team's context
Middle managers who understand and believe in the AI tool can bridge the gap. Those who are simply relaying executive enthusiasm cannot.
Remove Barriers Before Adding Motivation
Audit the actual barriers to adoption:
- Is the tool integrated into existing workflows or bolted on as an extra step?
- Does the training match the team's actual skill level?
- Are metrics adjusted for the learning curve?
- Is there a safe way to report problems with the AI tool?
- Do people know specifically how their role will change?
Address these before adding another communication campaign.
Measure Adoption at the Team Level
Organisation-wide adoption numbers hide variation. One team at 80% adoption and nine teams at 8% looks like 15% average. Understanding which teams adopt and which don't, and why, reveals the specific barriers rather than the aggregate number.
Create Feedback Loops That Reach the CEO
The CEO needs unfiltered information about what adoption actually looks like. Not the dashboard summary, not the project manager's optimistic update. Direct exposure to frontline experience, structured listening, and honest assessment of what's working and what isn't.
CEO championship of AI is necessary but not sufficient. The organisations that achieve genuine adoption are the ones that bridge the experience gap with structural support, not just enthusiasm. Champion the vision, but engineer the adoption.