I think the first line of an AI enablement budget should pay for managers’ time and training, ahead of another round of licences or company-wide prompt courses. Gallup’s latest research gives the clearest reason I have seen this year: the manager is where AI use turns into changed work, or stalls.
The number that should move the budget
Gallup and Silent Partners announced a joint venture on November 10, and the release carries findings from Gallup’s research. Two in three U.S. employees in organizations that have implemented AI say it has improved their personal productivity. Just 14% strongly agree it has transformed how work gets done. Silent Partners calls the distance between those numbers the AI adoption gap.
The finding I would put in front of a CFO is the next one. Employees whose managers actively support their team’s use of AI are 8.7 times more likely to strongly agree it has transformed how work gets done. It points at a specific role with a budget line HR controls.
Joe Daly, Global Managing Partner and Head of Strategic Ventures at Gallup, said: “As AI changes how work gets done, leaders need to give their teams clear direction and the support to put new capabilities to use.” I agree, and I would go further on the money. Direction and support take a manager’s hours. An AI budget that pays for tools and broad training without paying for those hours assumes they cost nothing.
What employees are signalling
Visier’s 2027 trends release, issued the same day, cites earlier Visier research in which 48% of employees admit to exaggerating their AI usage or expertise, and 59% say their employer either lacks a clear plan for how AI will affect jobs or they are unsure whether one exists. When half the workforce overstates its AI use and most cannot see a plan, a dashboard of logins and licence counts will flatter the rollout.
A manager who sits with the team and asks which tasks have changed sees what the dashboard misses. That conversation is also where an employee can ask the question the Gallup release says people are asking: what is expected of me now.
The boardroom shows the same blind spot from above. Deloitte Global’s survey of 806 board members and executives, which we covered in our report on the 16% with a workforce strategy for technology change, found 49% have limited or no oversight of how technology-transformation spending is split between infrastructure and workforce enablement. If directors cannot see the enablement line, nobody is asking whether managers are funded inside it.
The strongest objection
The best case against my view is that managers are a bottleneck. Employees adopt tools faster than their bosses, and routing enablement through managers slows the people who are already ahead. A second objection is about the evidence. The 8.7 times figure compares employees with and without that support, and teams with supportive managers may already be the teams most inclined to change how they work. A third is that Silent Partners sells training that starts with senior leaders, so a manager-first finding suits its offer.
I take all three seriously. On the first, an employee who is ahead keeps the tool. The budget question is where the next dollar goes, and the employee who is already ahead does not need it as much as the team whose manager has never redesigned a workflow. On the second, an association this large is still the best evidence an HR leader has on where change sticks, and the cost of acting on it is modest: manager time, which HR can measure and reverse. On the third, the figure is Gallup’s research, and the conclusion I draw from it does not require buying anyone’s program. Andrew Peters, founder and CEO of Silent Partners, said in the release: “And leaders have to go first, because people can’t follow you somewhere you haven’t been.” His point holds whether the training is bought or built in-house.
What I would fund
I made a related argument earlier, that executive AI usage is a poor proxy for workforce readiness. Manager support is a better proxy, and it can be budgeted. Concretely, I would ask for four things.
First, protected time: a fixed number of hours a month for each people manager to review with the team which tasks AI now handles and which it should. Second, training for managers on redesigning a workflow, separate from training on prompts. Third, one engagement survey item, asked every cycle, on whether the employee’s manager supports the team’s use of AI, so HR can see the measure Gallup used in its own workforce. Fourth, a line in the AI budget that shows enablement spending separately from tools, which is the split Deloitte’s respondents say they cannot see.
If the survey item rises and the share of employees who say their work has changed does not, HR will know within two cycles that the money should go elsewhere. That is a cheaper test than another year of licences.
Source: Gallup and Silent Partners, PR Newswire. Also drawn from Visier and Deloitte.
