Adoption is outrunning readiness
Kyndryl's 2026 People Readiness Report, a global study of 1,100 senior leaders, found that 57% of organizations have now embedded AI in core business processes. A year earlier that figure was 35%. Deployment is moving quickly and few boards are asking whether it can move faster.
The readiness picture moved the other way. Only 23% of leaders say their workforce is fully ready to use the AI they have deployed, a six point decline on the previous year. And 79% agree that the pace of AI will outrun their workforce, their governance and their operating models.
Read those two findings together and the problem states itself. The gap slowing most organizations is not a technology gap. It is a readiness gap, and it is widening while the technology spend accelerates.
The ROI problem is a people problem
McKinsey's 2026 State of AI survey found that 80% of employees report individual productivity gains from AI. Only 37% say it has contributed to company earnings.
That distance between personal productivity and enterprise value is the whole story. Deployment has become straightforward. Converting deployment into a number the board recognizes has not. Kyndryl found that only 11% of organizations had achieved both of their top two AI goals.
Individual gains do not aggregate on their own. They aggregate when the work itself is redesigned around the new capability, when handoffs between roles are rebuilt, and when the time saved is deliberately redirected rather than quietly absorbed. None of that is a software task.
Readiness is built, not bought
Gallup's July 2026 research makes the point sharply. Access to AI on its own did nothing for engagement, which held flat at 31%. Giving people the tool changed nothing measurable.
Three conditions changed the result:
- Frequent use. Occasional use produces no durable change in how work gets done.
- A clear implementation plan. Associated with a 15 point lift in engagement.
- Active manager support. Moves engagement from 30% to 48%.
Where all three conditions were present, engagement rose to 53%. The difference between a licence that sits unused and a workforce that performs differently is not the model. It is the plan and the manager.
What the organizations getting returns actually do
Kyndryl identifies a group it calls Pacesetters, the 9% of organizations converting AI investment into results. They do three things before they scale.
- They redesign roles around AI rather than adding AI to existing roles.
- They manage the change deliberately, as a programme with an owner and a measure.
- They build workforce readiness first, not alongside, and not afterwards.
These organizations are 1.5 times more likely to achieve AI related revenue growth. The pattern is consistent across the research. AI returns are produced by the operating model and by manager capability, not by tool licences.
What this means for people leaders
If your AI programme is reported on in licences issued, active users and pilots launched, you are measuring deployment. None of those figures tell you whether the work changed.
Three questions are more useful at this stage. Which specific roles have been redesigned around the capability, rather than handed an extra tool. Who owns the adoption plan as a deliverable with a date and a measure. And are your managers equipped to lead it, given that manager support is the single largest lever in the data.
The organizations treating adoption as a workforce transformation, rather than a software rollout, are the ones closing the gap between spend and return. That is a people strategy question, and it belongs on the leadership agenda before the next phase of scaling, not after it.
About FT Consulting Partners
FT Consulting Partners is a practitioner-led consulting firm. We partner with CHROs, CEOs and senior leaders on workforce and people transformation, change management and AI integration, built by people who have led transformation from inside organizations rather than advised from the outside.