Part 3 - Build a team that can actually use AI
AI adoption isn’t a technology problem. It’s a leadership problem.
Giving people access to AI does not mean they will know how to use it well. Some will move quickly. Others will avoid it. Different teams may create their own rules, standards and ways of working.
Leaders need to create clarity around how AI should be used, how people will learn and where the boundaries sit.
Decision 7: How your team works with AI
Without clear expectations, adoption becomes inconsistent. Teams need to understand:
When AI should be used
How its outputs should be checked
What information must never be entered
When human judgement is required
Who remains accountable
Clarity turns AI from an individual experiment into a consistent way of working.
Where leaders get it wrong
Assuming everyone will adopt AI in the same way
Leaving teams to create their own standards
Introducing tools without defining new ways of working
From the seat
The AI workflows I built did not stay with me, and that was the point. While they lived only in my hands, they were a party trick. The number moved once the team could run them without me. Capability that lives with one person is not capability. It is a risk you have not named yet.
Make the call
Agree on three practical principles for AI use across the organisation. For example:
● We use AI to improve work, not replace critical thinking.
● We check AI generated outputs before sharing them.
● We protect customer and business information at all times.
Ask yourself
If every team created its own rules for AI, how much inconsistency and risk would that create?
Decision 8: How your team learns
Most teams are handed AI and left to it. 72% of Australian small and medium businesses have no plans to offer AI training. Access is not capability.
AI capability will not come from a single workshop. The organisations that benefit most will not necessarily be those that know the most today. They will be the ones that keep learning.
Capability grows through:
Regular use
Shared examples
Open discussion
Experimentation
Reflection on what worked and what did not
Where leaders get it wrong
Treating AI learning as something you do once
Expecting people to learn in their own time
Rewarding perfection instead of experimentation
Make the call
Create one regular opportunity for your team to learn together. It could be:
Five minutes in a team meeting
A monthly AI discussion
A demonstration of a useful application
A shared review of what worked and what did not
Small, consistent learning moments will have more impact than one large training session.
Ask yourself
In twelve months, will your team see AI as another tool they were given, or a capability they built together?
Decision 9: Responsible AI
Responsible AI gives people the boundaries they need to move with confidence.
Leaders must set expectations around:
Privacy
Security
Confidential information
Checking AI generated work
Accountability for decisions
Transparency with customers
Responsible AI is not someone else’s responsibility. It is part of how the organisation chooses to operate.
Where leaders get it wrong
Assuming everyone knows what responsible AI means
Overlooking privacy and confidentiality risks
Treating AI governance as a technical issue
Make the call
Ask your leadership team: if a customer asked how we use AI, could we answer clearly and confidently? If the answer is no, your organisation does not yet have a clear position.
Ask yourself
What reputation do you want your organisation to build as AI becomes a bigger part of how you work?
Decision 10: Your first move
It is easy to spend months exploring tools, comparing options and discussing possibilities. Progress begins when a leader chooses what matters and acts.
You do not need to solve everything at once. You need one meaningful next step that:
Has a clear purpose
Creates momentum
Has an owner
Has a date
Produces something you can learn from
Where leaders get it wrong
Waiting until every question has been answered
Trying to transform everything at once
Confusing planning with progress
Make the call
Review the ten decisions and identify the one that could have the greatest impact over the next 90 days. Choose one action. Assign an owner. Commit to a date. Then take the first step.
Don’t try to do all ten this week. Pick one.
These decisions are not a checklist to complete once. They are conversations worth returning to as AI becomes a bigger part of how your organisation works.
Choose the decision that matters most now. Take one action. Build from there.
You did not need to lead the technology. You needed to lead the people using it. That was the whole job.
