Employees with supportive managers are far more likely to use AI. Here's what AI training for managers should cover, and how to run it so teams don't drift back.
TL;DR
When AI programmes stall, attention usually goes to the tools or the training content. The more useful place to look is often the middle of the organisation. Managers decide what gets prioritised this week, what counts as good work, and whether trying something new is safe. If they're unsure about AI, their teams will be too.
This piece covers why managers need their own AI training, what it should include, and how to run it.
Gallup's 2026 research on AI and workplace productivity puts numbers on something many leaders sense. Manager support is one of the two conditions most strongly associated with frequent AI use, alongside how well AI fits into existing workflows.
The gap is large. In organisations adopting AI, 78% of employees who strongly agree their manager supports AI use say they use it frequently, compared with 44% of those who don't strongly agree. Those with supportive managers are also 9.3 times as likely to strongly agree that AI has transformed how work gets done.
And yet only 36% of employees in AI-integrating organisations strongly agree their manager actively supports their team's use of AI. Gallup's tracking also shows frequent use among managers rising from 15% in 2023 to 36% in 2026, while among senior leaders it rose from 17% to 51%. Managers are being asked to lead something many of them aren't yet doing much themselves.
Managers can't credibly encourage something they don't do. Start with hands-on use on management tasks: preparing for one-to-ones, summarising project updates, drafting difficult messages, writing up performance notes for their own reference, and planning team priorities. The goal is regular, comfortable use within a few weeks.
Gallup describes effective manager support in four behaviours: actively supporting AI use at team level, modelling it, answering employees' questions, and showing how AI connects to employees' daily work. That's a good structure for this module.
In practice, it means a manager who asks in team meetings what people have tried, shares their own examples including the failures, knows the rules well enough to answer "am I allowed to put this in?", and helps each person find one or two tasks where AI helps.
This is where measurable change happens. Managers need to look at their team's recurring work and identify which steps AI could speed up, which need a human check, and which might disappear. Then they need to change the process, set the expectation, and allow time for the new way to bed in, including the period where it's slower.
People tend to ask their manager, rather than the CEO, whether AI will affect their job. Managers need a clear, honest answer that's consistent with the organisation's position, and the confidence to say "I don't know yet" when that's the truth. At Praxes, our view is that AI-native work should make teams more capable, and that's a much easier conversation to have when it's also the organisation's intent.
Managers review their team's work, which increasingly includes AI-assisted work. They need to recognise when output has been checked and when it's been pasted in unread. Harvard Business Review argues that as AI makes polished work easier to produce, judgement becomes the scarce skill. Managers are the people best placed to build it in their teams.
Some team members will already be using AI tools the organisation hasn't approved. Managers need to know how to respond without driving the behaviour underground. Our guide to shadow AI covers how to manage it without a blanket ban.
Train managers first. Run the manager session a week or two before team sessions. Managers who've already tried the tools can answer questions and reinforce the training afterwards.
Keep it practical and short. A half-day session covers personal use, the four championing behaviours, and a first pass at workflow redesign for each manager's team. Ask each manager to bring one recurring team process to work on.
Give them a simple routine. A five-minute AI slot in the weekly team meeting, where someone shares what they tried, does more for adoption than any poster campaign. Managers need permission and a format to make it a habit.
Build in check-ins. Two and four weeks after the session, ask managers what changed in their team. The Kirkpatrick model lists manager check-ins as one of the support mechanisms that help learning turn into behaviour, and that applies to managers as learners too. Our guide to measuring AI training covers how to track it.
Use champions as support. Not every manager will be an enthusiast. Pair less confident managers with one of the champions Atlassian's workshop play recommends appointing in each team: someone who uses AI well and can carry some of the hands-on support.
For individual managers who want structured learning, Microsoft and LinkedIn publish an AI for Managers learning path covering practical uses such as making meetings more effective and giving feedback. It's a reasonable foundation.
What it can't do is connect AI to your team's specific work, your rules, and your organisation's position on AI and jobs. Pair it with an in-house session that does.
Senior leaders need a different conversation, focused on strategy, investment, and governance. Our guide to AI training for executives covers that level.
Because manager support is strongly linked to whether employees use AI. Gallup found 78% of employees with actively supportive managers use AI frequently, compared with 44% of those without that support.
Their own use on management tasks, how to champion AI in their team, how to redesign their team's workflows, how to talk honestly about AI and jobs, how to judge AI-assisted work, and how to handle unapproved tools.
Microsoft and LinkedIn's AI for Managers learning path is a solid individual foundation. For lasting change, combine it with an in-house session built around your team's work and your organisation's rules.
Yes, ideally a week or two before. Managers who've used the tools can answer questions, model use, and reinforce the training when their team comes back to work.
Gallup lists four behaviours: actively supporting AI use, modelling it, answering employees' questions, and showing how AI connects to their daily work. A short weekly slot where people share what they tried is an easy way to start.
If you'd like to know how ready your managers and teams are before planning training, our free AI Readiness Diagnostic gives you a scored view across five dimensions, including people and process.
Ready to move from reading about AI to getting your organisation ready for it? Start with the diagnostic.
Take the AI Readiness Diagnostic