Why AI adoption lives or dies with your line managers

As regular readers will know, we take a look at Gallup’s State of the Global Workplace report every year to see what it tells us about the world of work, employee engagement and, inevitably, why everything isn't quite going according to plan. We looked at the 2026 report in some detail back in April, but it raised a particularly interesting collision between two subjects we've recently been talking about (AI adoption and manager communication) so we decided an extra instalment was in order.

Mainly because organisations everywhere are currently trying to work out what AI means for them.

Copilots are being launched. Licences are being bought. Policies are being written. Training programmes are being created. Senior leaders are attending conferences and returning with the words “agentic AI” written excitedly in their notebooks.

And somewhere in the middle of all this sits the humble line manager.

Already managing performance, workload, wellbeing, hybrid working, organisational change, development conversations, restructures, new systems, employee expectations, and somebody called Gary who keeps microwaving fish in the communal kitchen.

Now we're adding:

“Oh, and could you make sure your team embraces AI?”

Good luck, Gary's manager.

But Gallup's latest research suggests this isn't a minor implementation detail.

It could be the difference between AI becoming genuinely useful and your expensive new technology becoming another icon employees occasionally look at before opening Google.

Managers are the missing link

Gallup's State of the Global Workplace 2026 makes the scale of the AI opportunity pretty clear. But it also makes something else clear:

“Managers play a critical role in meaningful AI adoption.”

That's Gallup's wording, not ours.

Based on its Q1 2026 US workforce research, Gallup found that the two biggest drivers of frequent organisational AI use were integration with existing work systems and manager-led AI adoption.

When employees strongly agreed that their manager actively supported their team's use of AI, 79% were frequent AI users, compared with 46% where that manager support wasn't present.

It gets considerably more interesting.

Among employees working for organisations investing in AI, those who strongly agreed that their manager actively supported their team's use of AI were nearly nine times more likely to strongly agree that AI had transformed how work gets done.

They were also over seven times more likely to strongly agree that AI gave them more opportunities to do what they do best every day.

Those aren't subtle differences.

That's not “Managers make things approximately 4% nicer.” That's an enormous gulf in how employees experience and value AI.

And it points towards something organisations risk missing in the excitement around technology.

AI adoption is a people change programme.

You can call it digital transformation if you like. You can call it an AI enablement programme. You can give it an exciting name and create a logo featuring some interconnected hexagons.

But ultimately, you're asking human beings to change the way they work.

And that makes it a change communications challenge.

Unfortunately, managers are knackered

Here's where things become slightly awkward.

At precisely the moment organisations desperately need managers to help employees navigate perhaps the biggest workplace technology shift in decades, managers aren't doing particularly well themselves.

Gallup found that global manager engagement fell from 31% in 2022 to just 22% in 2025.

Individual contributor engagement, meanwhile, moved from 20% to 19% over the same period. Gallup therefore concludes that declining manager engagement accounts for much of the recent overall deterioration in employee engagement.

The largest annual fall came between 2024 and 2025 alone, when manager engagement dropped from 27% to 22%.

Gallup rather neatly describes what's disappearing as the “engagement premium” traditionally associated with being a manager. Managers are increasingly becoming almost as disengaged as the people they're supposed to be engaging.

Excellent timing.

Gallup's wider research estimates that managers account for 70% of the variance in team engagement, so this isn't simply a manager wellbeing issue tucked away for HR to worry about.

Managers are the layer through which an enormous amount of organisational reality gets translated.

Strategy becomes priorities.

Change becomes action.

Policy becomes behaviour.

And AI becomes either “Here's how this could genuinely make our team's work better.” or “Apparently we're supposed to use Copilot now.”

Those are two very different AI rollouts.

Don't just give managers a toolkit

This is where Internal Comms needs to be careful.

We love a manager toolkit.

• Talking points!
• FAQs!
• Slides!
• Key messages!
• Printable conversation guide!

Lovely.

Pop everything on the intranet, send managers an email saying PLEASE CASCADE TO YOUR TEAMS, and our work here is done.

Except it isn't.

If managers are central to AI adoption, they need more than communication materials. They need confidence.

Imagine being asked questions like:

• “Will AI eventually replace my job?”
• “Am I allowed to put this customer information into it?”
• “Are you measuring whether I use it?”
• “Does using AI mean I'm cheating?”
• “Will my role change?”
• “What happens if it gives me the wrong answer?”
• “Why should I use this when my current process works perfectly well?”
• “Are you eventually going to need fewer people?”

Those aren't questions you answer with Slide 26 of the AI Launch Deck.

Managers need to understand the organisation's position well enough to have an actual conversation. That means bringing them into the change early.

• Explain why the organisation is adopting AI, not simply which button to press.
• Give them practical examples relevant to their teams.
• Be clear about guardrails and acceptable use.
• Tell them what you know.
• Tell them what hasn't been decided.
• Give them somewhere to escalate questions they cannot answer.
• And, critically, let managers ask their own difficult questions before employees start asking them.

Because nothing builds confidence in an organisational transformation quite like:

“Good question. I only found out about this from the same email you did.”

Make manager briefings safe

This is also where the format of manager communication matters.

If you're introducing AI into the workplace, there will be questions people don't necessarily want to ask publicly.

Managers themselves may feel embarrassed admitting they don't understand something. They may worry that questioning the AI strategy makes them look resistant to change. They may be concerned about their own role.

Or perhaps somebody has been talking confidently about large language models for 45 minutes and everyone in the room is now too frightened to admit they don't know what an LLM is. (Let’s be honest; some of them are still picturing the T-1000 from The Terminator.)

Create space for honesty.

Live manager briefings can work extremely well, particularly if they're genuinely interactive rather than another presentation being inflicted upon people.

Tools such as Vevox audience polling can also allow anonymous Q&A, giving managers a safer way to surface concerns, challenge assumptions and ask the questions they'd rather not attach their name to.

And don't be frightened by awkward questions. They're useful.

If 40 managers independently ask:

“How do I know what information employees can safely enter into this?”

you haven't discovered a manager resistance problem. You've discovered a clarity problem. And that's valuable intelligence. Use it!

Give them something worth cascading

Then we get to the actual communications.

Your managers shouldn't need a degree in Corporate Translation Studies to explain your AI programme.

• Give them a simple story.
• Why are we doing this?
• What problems are we trying to solve?
• What are we introducing?
• What does it mean for employees now?
• What might change later?
• What are the boundaries?
• Where can people experiment safely?
• Where can they learn?
• Where can they get help?
• And what do we want managers to do?

That's the basis of a strong manager communication pack.

Not 64 slides. Not the entire AI strategy. Not a document containing every answer Legal has ever approved.

Give managers the core narrative, then supporting material they can reach for when they need it.

Real examples will help enormously too. “AI will unlock productivity” means very little. “Here's how Sarah's team is using it to turn a two-hour weekly reporting task into a 20-minute one” means considerably more.

This is where good internal comms campaigns earn their keep. Instead of one enormous AI launch followed by radio silence, build adoption over time.

• Show use cases.
• Tell employee stories.
• Use short videos.
• Create demonstrations.
• Share mistakes and lessons.
• Celebrate useful experimentation rather than simply usage numbers.
• Answer new questions as they emerge.
• Give managers fresh material for team conversations.
• And communicate progress.

The objective isn't: EVERYONE MUST USE AI BY FRIDAY.

It's helping people understand where AI genuinely improves their work and giving them enough confidence to try it.

Map the humans, not just the technology

AI rollouts also tend to produce extraordinarily complicated stakeholder landscapes.

IT owns one bit. HR owns another. Legal has opinions. Information Security has considerably stronger opinions. The Executive Team wants progress. Internal Comms wants clarity. Employees want to know whether robots are coming for their jobs. And managers are sitting somewhere in the middle attempting to reconcile all of it.

This is exactly why stakeholder mapping matters. Before launching the communications campaign, work out who can influence adoption, who employees trust, who needs deeper information and where resistance or confusion might appear.

Our free Stakeholder Management Plan is designed to help structure exactly that kind of thinking.

And don't automatically categorise managers as one homogenous audience called “Managers”. A senior manager leading 150 knowledge workers may have completely different AI opportunities and concerns from a frontline supervisor leading a team of warehouse employees.

Their teams may have different access. Different use cases. Different risks. Different levels of confidence. Different fears.

Your communication should recognise that.

Personalisation doesn't mean producing 97 versions of every email until the Internal Comms team collectively resigns. It means understanding where differences actually matter.

Support, don't just recruit

There is another statistic in Gallup's report that should make organisations pay attention.

Despite the relationship between manager support and successful AI adoption, less than a third of US employees in organisations already implementing AI strongly agreed that their manager actively supported their team's use of it.

Gallup found a similarly low figure in Germany: 21%. So, there's a fairly substantial gap between managers matter enormously and managers currently feel equipped to actively champion this.

The answer cannot simply be to demand more from them.

Gallup makes an interesting observation later in the report: effective people management is itself a skill, yet many managers haven't received the training required to coach teams effectively. It suggests AI might eventually help managers themselves by providing personalised, real-time management advice.

There's a nice irony there.

The people we're asking to help everyone else adopt AI may themselves be among the groups who could benefit most from it.

So, show managers that too.

• How could AI help them prepare for meetings?
• Summarise information?
• Develop questions for one-to-ones?
• Analyse themes?
• Reduce administrative work?
• Create first drafts?
• Spend less time formatting PowerPoint and more time actually managing humans?

Don't just recruit managers as unpaid sales representatives for the AI programme. Help them experience the benefit.

Advocacy tends to become considerably easier when somebody actually believes in the thing they're advocating. (Revolutionary, innit?)

This is about trust, too

We've previously written about talking about AI without scaring your workforce, because employee concerns about AI aren't irrational.

• Jobs may change.
• Skills will change.
• Some work will disappear.
• New work will emerge.
• There are legitimate questions around privacy, security, accuracy, ethics and job security.

Pretending otherwise won't build confidence.

And that's another reason line managers matter so much. Employees don't experience organisational change solely through corporate announcements.

They experience it through conversations.

• What does my manager think?
• Are they using it?
• Do they understand it?
• Can they answer my questions?
• Are they worried too?
• Am I safe admitting that I'm struggling?
• Can I experiment without being punished for getting something wrong?

Gallup's findings reinforce the argument we made in our earlier piece on Manager Communication: The Most Important Channel You're Probably Ignoring.

Your managers aren't simply another audience on the comms plan. They're a communication channel in their own right.

And when the change is as potentially personal as AI, they're one of your most important.

Your AI strategy needs a manager strategy

The State of the Global Workplace 2026 tells a fascinating story.

Global employee engagement fell to 20% in 2025, its lowest level since 2020. Gallup estimates that low engagement cost the global economy around $10 trillion in lost productivity during the year.

At the same time, organisations are investing heavily in technology they hope will produce enormous productivity gains.

There's a fairly obvious lesson sitting between those two facts. Technology alone doesn't transform organisations. People do.

And Gallup's AI findings suggest line managers are going to play an enormous role in determining whether employees understand, trust and actually use these tools.

So, before your AI launch, ask:

• Do managers understand why we're doing this?
• Have they had the opportunity to challenge it?
• Can they answer the obvious questions?
• Do they know what good AI use looks like for their teams?
• Are the guardrails clear?
• Can they demonstrate useful applications themselves?
• Do they know where to send questions they can't answer?
• And perhaps most importantly:

Have we actually supported them, or have we just given them another thing to cascade?

Because your organisation can buy the best AI technology available. You can build a beautiful launch campaign. You can produce videos. Create guides. Design a spectacular intranet hub.

You can even create an AI mascot called Copilot Colin if things have really got out of hand.

But ultimately, employees will decide whether AI belongs in their working lives through everyday experience. And a huge part of that experience will be shaped by the person they work for.

• So yes, invest in the technology.
• Invest in the change communications.
• Build a proper internal comms campaign.
• Map your stakeholders.
• Create genuine opportunities for questions.

But don't forget the people standing between your AI strategy and the employees expected to make it real.

Because if Gallup's latest research tells us anything, it's this:

Your AI rollout might have been designed in the boardroom. But it's going to live or die in the team meeting.

Rolling out AI? Let’s talk

We help organisations plan the people side of AI rollouts, from manager briefings to full change comms campaigns. Book a free discovery call and tell us where you are.