Responsibility sits in everyday calls about what is acceptable, where to encourage experimentation, and where to hold back. But when governance is still forming, those calls are made individually. One division approves AI-assisted client recommendations. Another prohibits them. Neither knows the other's position. The intent may be the same. The outcome is inconsistency, and in a connected labour market, inconsistency is noticed fast.
The longer this continues, the more the risk compounds. Not dramatically. Quietly. One uncoordinated decision at a time. One team building confidence in a tool that another team is quietly told to avoid. Over months, what looks like progress becomes fragmentation. Standards drift. Effort is duplicated, and fragmentation, once embedded in how an organization operates, is very difficult to reverse.
This is not a story about reluctance. More than 90% of Caribbean leaders are open, positive, or cautiously hopeful about AI. The dominant posture across the region is willingness. What the data shows is that willingness without shared structure produces motion without direction. People are making decisions. They are making them alone, without common ground to stand on.
There is also a gap in how learning moves. Leaders are experimenting and adjusting, finding ways to apply AI in their own environments. But that learning develops within teams, sometimes within individuals, and then stalls. Across sectors, similar problems are being solved in parallel without connection. The appetite for learning is evident across the region. The pathways that would let it travel and build are not.
These are not separate problems. They are expressions of the same condition: organizations doing more with AI than their leadership systems were designed to guide. And the three-speed problem tells us that the gap is not closing on its own. It is widening.
What the research points toward is not a set of technical fixes. It is a shift in how leaders treat AI within their organizations. The question is no longer whether AI will be used. It is whether leaders can build enough shared understanding, visible governance, and connected learning to guide what is already underway.