When I ask workshop participants what enables their learning at work, the most common answer isn’t training programs or learning platforms. It’s their manager.

Managers create the conditions for learning—or destroy them. They allocate time, model behaviour, provide psychological safety, and reinforce new skills. No amount of excellent training overcomes a manager who doesn’t support it.

If you’re managing a team through AI adoption, here’s how to create an environment where learning actually happens.

Why Your Role Matters

Research consistently shows that manager behaviour is among the strongest predictors of training transfer—whether learning gets applied in actual work. Studies cited by Harvard Business Review suggest that manager support may be the single most important factor in whether employees apply new skills.

Managers influence learning through:

Time allocation. You control workload and priorities. If there’s no time for learning, there’s no learning.

Psychological safety. People take risks and try new things when they feel safe. You determine that safety.

Expectations. What you expect and communicate shapes behaviour. If you don’t expect AI fluency, you won’t get it.

Modelling. If you use AI effectively, your team sees it’s expected and valued. If you don’t, they get a different message.

Reinforcement. Recognising and rewarding new behaviours makes them stick. Ignoring them lets them fade.

Every manager action either supports or undermines AI capability development. There’s no neutral position.

Creating Time for Learning

The first practical challenge is time. Your team is busy. So are you. Learning feels like something extra.

But without time, nothing else matters. Here’s how to create space:

Protect Learning Time

Block dedicated time for AI experimentation and learning. This means:

• Calendar blocks that don’t get overrun
• Permission to decline meetings during learning time
• Expectation that this time is legitimate work

How much? At minimum, a few hours per week during active learning phases. More is better.

Reduce Competing Priorities

If you protect learning time but maintain impossible workloads, people will skip learning to catch up. Something has to give.

Review priorities. What can be deprioritised or eliminated to create genuine capacity for learning?

Integrate Learning into Work

Rather than treating learning as separate from work, integrate it:

• “Try using AI for this report and see if it helps”
• “Let’s use AI to prepare for this meeting”
• “Before doing this manually, experiment with AI approaches”

This makes learning part of work, not competition with work.

Building Psychological Safety

Learning requires taking risks. Trying new approaches. Making mistakes. Looking incompetent while developing competence.

If people don’t feel safe doing these things, they won’t. They’ll stick with what they know.

Normalise Experimentation

Make clear that trying new things—and sometimes failing—is expected:

• “I want you to experiment with AI. Some attempts won’t work. That’s fine.”
• “Tell me what you tried and what you learned, not just what succeeded.”
• “I’d rather you try something new and struggle than play it safe.”

Share Your Own Learning

Let your team see you learning, including your struggles:

• “I tried using AI for this and it didn’t work. Here’s what happened.”
• “I’m still figuring this out too. Let’s learn together.”
• “I made a mistake relying on AI output without checking. Here’s what I learned.”

Vulnerability from leaders creates safety for others.

Respond Constructively to Mistakes

How you respond to mistakes matters enormously. If someone tries AI and it doesn’t work:

• Explore what happened with curiosity, not blame
• Focus on learning, not fault
• Help them figure out what to try next
• Thank them for experimenting

One punitive response undoes months of safety-building.

Setting Clear Expectations

People need to know what’s expected. Vague hopes don’t drive behaviour.

Be Specific About AI Use

Clarify what you expect:

• “I expect everyone to try AI for at least one task this week.”
• “When drafting documents, explore whether AI can help before starting from scratch.”
• “Share one AI tip or learning in our team meeting each week.”

Specific expectations can be met. Vague ones can’t.

Connect to Performance

Include AI capability in performance discussions:

• Is developing AI fluency part of their goals?
• Do performance conversations include AI learning progress?
• Does recognition include AI skill development?

If AI isn’t part of performance, it’s not really expected.

Communicate the Why

Expectations land better with rationale:

• “AI fluency is becoming essential in our field. I want you to be ahead of the curve.”
• “This helps you personally and helps the team deliver better work.”
• “The skills you develop now will serve you throughout your career.”

People engage more when they understand purpose.

Modelling the Behaviour

People watch what you do more than what you say. Model the AI use you want to see.

Use AI Visibly

Let your team see you using AI:

• “I used AI to draft this—let me show you the process.”
• “Here’s how I’m using AI to prepare for our strategy session.”
• “I’m experimenting with AI for [specific task]. I’ll share what I learn.”

Share Your Learning Journey

Don’t just show polished results. Share the process:

• What you’re trying to figure out
• What hasn’t worked
• What you’re learning along the way

This normalises learning as ongoing, not a destination.

Ask for AI Ideas

Invite reverse mentoring:

• “I saw you use AI effectively for that project. Can you show me how?”
• “Anyone found a good AI approach for [challenge]?”
• “What are you all learning that I should try?”

This signals that learning flows in all directions.

Reinforcing New Behaviours

Behaviours that get reinforced persist. Behaviours that don’t fade.

Recognise AI Learning

Notice and acknowledge when people develop AI capabilities:

• “I saw you used AI to speed up that analysis. Nice work figuring that out.”
• “Thanks for sharing that AI tip with the team.”
• “Your AI skills have really improved this quarter.”

Create Sharing Opportunities

Build team mechanisms for AI learning:

• Regular sharing in team meetings
• Slack/Teams channel for AI tips
• Buddy systems for AI learning support
• Team celebration of AI wins

Social reinforcement often matters more than formal recognition.

Connect to Outcomes

Help people see how AI learning contributes to outcomes:

• “That AI approach saved significant time. The client noticed.”
• “The quality of this analysis—partly because you used AI well—strengthened our recommendation.”
• “Your AI skills made this project possible.”

Outcome connection makes learning feel worthwhile.

Handling Resistance

Some team members may resist AI adoption. Understanding the resistance helps you address it.

Understand the Concern

Resistance usually has roots:

• Fear about job security
• Concern about looking incompetent
• Scepticism about AI value
• Discomfort with change
• Practical barriers to access or time

Listen to understand before trying to convince.

Address Legitimate Concerns

Some concerns are legitimate and deserve direct engagement:

• If job security is a worry, provide honest reassurance (if you can)
• If access is a barrier, solve the access problem
• If time is the issue, address time allocation

Respect Pace Differences

People adopt change at different speeds. Early adopters will embrace AI immediately. Others need more time.

Allow pace differences within a reasonable range:

• Set minimum expectations everyone must meet
• Allow faster adopters to go further
• Provide extra support for slower adopters
• Don’t let anyone opt out entirely

Know When to Escalate

Persistent resistance despite support eventually becomes a performance issue. If someone refuses to develop capabilities that are genuinely required for their role, that’s a conversation about performance expectations.

This should be rare. Most resistance dissolves with proper support.

Practical Manager Actions

Here’s a weekly checklist for managers supporting AI learning:

Weekly:

• Check in on AI learning progress with each team member
• Share something you’re learning about AI
• Create or protect time for team AI experimentation
• Recognise AI learning or application you’ve observed

Monthly:

• Review whether AI learning is progressing across the team
• Identify anyone who needs additional support
• Share team AI wins more broadly
• Assess whether time and resources are adequate

Quarterly:

• Include AI capability in performance discussions
• Evaluate overall team AI readiness
• Identify next-level AI development needs
• Adjust expectations based on progress

The Multiplier Effect

When managers effectively support AI learning, the impact multiplies. Each team member who develops capability can help others. Effective practices spread. The organisation’s overall AI fluency accelerates.

When managers don’t support learning, even excellent training programs fail. People attend, learn something, return to work, and revert to old patterns. The investment is wasted.

You’re the critical variable. The good news is that the actions required aren’t complex. Time protection, psychological safety, clear expectations, modelling, and reinforcement—these are basic management behaviours applied to AI learning.

The organisations that build AI capability fastest will be those where managers take this role seriously.

Your team is waiting for your leadership. Give them what they need.