You bought the licences. You sent the email announcing them. Maybe you ran a lunchtime demo. Three months later, AI adoption in your team looks like this: two enthusiasts use it every day, a few people tried it once, and everyone else has quietly gone back to doing things the old way.
That’s normal, and it’s fixable. This article explains why teams stall, why fear and confidence matter far more than features, and a 30-day test that turns AI from a novelty into a habit.
The adoption gap is not about the tools
Access stopped being the problem some time ago. The AI Economy Ireland 2026 report from Trinity College Dublin and Microsoft Ireland found that 92% of Irish organisations use or plan to use AI. The tools are cheap, and most of them are already inside the software you pay for.
Regular use is a different story. BCG’s AI at Work 2025 survey of more than 10,600 workers in 11 countries found that only 51% of frontline employees were regular AI users, a figure that had stopped growing. Managers were well ahead of them, at 78%.
So the tool is there and the people aren’t using it. Swapping ChatGPT for Copilot, or Copilot for Gemini, won’t change that. Something else is going on.
Fear is the real blocker
When someone in your team doesn’t use AI, the reason is usually one of these:
- Fear for their job. If the message from the top is “AI will make us more efficient”, plenty of people hear “AI will make some of us unnecessary”. They’re not going to rush to help.
- Fear of looking foolish. Trying something new in front of colleagues, and getting a daft answer, is uncomfortable. Easier not to try.
- Fear of getting it wrong. If nobody has said what’s allowed, the cautious choice is to do nothing. Your most conscientious people are often the most hesitant for exactly this reason.
- Fear of being judged for using it. Microsoft and LinkedIn’s 2024 Work Trend Index found that 52% of people using AI at work were reluctant to admit using it for their most important tasks. They worry it makes them look replaceable, or lazy.
The hesitation isn’t spread evenly either. The Trinity College and Microsoft research found that 70% of women hesitate to use AI at work, compared with 52% of men. If your rollout plan assumes everyone starts from the same place, it’s going to miss a lot of people.
These fears fade when a manager says clearly what’s allowed, what isn’t, and what AI means for people’s jobs, and then behaves consistently with that.
The confidence problem
Even when fear is handled, confidence often isn’t. Here’s the pattern I see most. Someone opens an AI tool, types a vague one-line request, and gets back a polite, generic paragraph. They decide it’s a clever toy that doesn’t understand their work, and they never come back.
The tool did what it was asked. A vague request gets a vague answer. Nobody had shown them that the useful version involves context, examples and a back-and-forth, much like briefing a new colleague.
Training closes this gap, but only if it’s the right kind. The BCG research found that only 36% of employees felt adequately trained in AI, and that people with at least 5 hours of training, especially in person with coaching, were much more likely to become regular users. In Ireland, SMEs were more than twice as likely as large firms to have no formal AI training at all (15% against 6%), according to the Trinity College and Microsoft report.
Confidence comes from a few hours working on your own documents, with someone beside you when it goes wrong.
What actually works
From what I’ve seen work in small teams, these are the things that move the needle.
Give clear permission and clear rules. A one-page policy that says which tools are approved and what data stays out of them removes most of the “am I allowed?” hesitation. Our shadow AI article covers what to put in it.
Say what AI means for jobs. If you intend to use the time saved for more clients or shorter Fridays, say so. If you don’t know yet, say that too. Silence gets filled with the worst interpretation.
Train on real work. Skip the generic examples. Use the proposal they’re writing this week, the report they dread every month, the email they answer 20 times a week. Our Getting More From AI course is built around exactly this, with one rule throughout: the AI drafts, you decide.
Lead from the front. If the owner or manager uses AI visibly, shares what worked and admits what didn’t, the team follows. If they delegate “the AI thing” to the youngest person in the office, the team draws its own conclusions.
Find your champions and give them time. Every team has one or two people who’ve gone further than the rest. Give them an hour a week to help colleagues. Peer help is less intimidating than a trainer and far cheaper than a consultant. (Yes, I’m a consultant. I’m telling you this anyway.)
The 30-day habit test
Habits form through repetition on things that matter. So here’s a simple test to run with your team.
- Each person picks one task. Something they do at least weekly that involves writing, summarising, reorganising or researching. Their choice, so they care about the result.
- Time it once the old way. Write the number down.
- For 30 days, use AI on that task every single time. No exceptions, even when it feels slower. The first week usually is.
- Keep a short log. Date, time taken, and one line on what worked or didn’t. A shared spreadsheet is plenty.
- Meet at day 30. Each person shares their before-and-after time and one thing they learned about getting better results.
Some tasks won’t improve, and that’s useful to know. Most will, and by day 30 the people doing them will have a habit, plus the evidence that convinces the sceptics in the room. Measured results from colleagues persuade people far more than any demo does.
If you’d like structured help, our AI training for teams is designed around this kind of practical, task-based approach. If prompting is the sticking point, Master AI Communication is a one-day workshop focused on getting professional-quality results from AI.