The Problem
AI transformation is 99% change management and 1% technology. So why is your IT team running it?
I've watched this play out in boardrooms across New Zealand. The CEO announces AI is a priority. The exec team nods. Then everyone looks at the CIO. And in that moment, the transformation is already dead.
The mistake isn't capability. It's framing. You called the biggest cultural shift of our lifetime a technology project. AI isn't a technology change. It's a people change. It's about humans understanding that the role they were hired for three years ago won't exist in the same form two years from now. And nobody in IT is trained to have that conversation.
Your competitors figured this out six months ago. They're not waiting for IT to approve a tool. They're educating their teams and seeing what's possible when failure is okay.
When was the last time someone on your team tried a new AI tool without asking permission first? If you can't remember, that's your answer.
What IT Holding You Back Looks Like
- •Every tool requires a security review that takes months
- •Free trials need procurement approval and three levels of sign-off
- •Someone mentions ChatGPT and ten people look nervous
- •Your AI policy is 34 pages long and nobody's read it
- •"Governance" delays decisions, not enables them
Your team stopped asking because they know the answer is no.
The Shadow AI Problem
Some employees are using AI quietly, getting work done faster, then filling the extra time looking busy doing tasks that don't matter. Because they don't know if it's allowed. That's shadow AI. And it's slowing you down more than the IT department.
This Isn't IT's Job. It's Yours.
Here's the uncomfortable truth. Your team isn't waiting for IT to give them access. They're waiting for you.
Every time you defer to the CIO, they notice. Every time you say "let's wait for the security review," they hear "not yet." Every time you frame AI as a technology decision, you give yourself an out. Someone else to blame when it stalls.
You set the temperature. And right now, you're cold as ice.
Your hesitation isn't neutral. It's permission for everyone else to stay still. That middle manager who's been dragging their feet? They're reading your energy. That team who keeps finding reasons not to try new tools? They learned it from watching you.
You might think you're being measured. Strategic. Waiting for the right moment. But your employees don't see that. They see a leader who hasn't committed. A leader who's still hedging. A leader who talks about AI in quarterly updates but hasn't stood up and said "this is happening and I'm leading it."
That's not strategy. That's avoidance.
The hardest conversation you need to have isn't with your CIO or with the board. It's with yourself. Are you actually ready to lead this?
The Leaders Who Stopped Waiting
Nobody has cracked the perfect AI transformation. Anyone telling you otherwise is selling something. But some leaders are ahead. Not because they figured it all out. Because they started before they were ready.
Case Study: Paul Pritchard, CEO of Overdose
Paul stood in front of hundreds of employees and said what most leaders won't. "Your jobs are going to change. The tools you rely on are being replaced. Some of what you do today won't exist in twelve months." Then he told them he was going to retrain every one of them to be ready.
No softening. Just truth.
This shifted the culture within the organisation. People started to try, experiment and share AI wins. They learned from each other. They improved their capabilities in house.
What Leaders Like Paul Have Done — And What You Should Do Next
Tell the Truth
Your team already knows change is coming. What they don't know is if you're leading them through it. Have the conversation. This week. Say it out loud.
Rewrite Your Values
Other companies have done this before touching a single tool. Courage. Curiosity. Momentum. Words that give permission to experiment. If your culture punishes failure, no technology will save you.
Go First
Use AI yourself before you ask anyone else to. Your team is watching. When they see you using it daily — not talking about it — resistance dissolves.
Find Your Champions
They're already in your business. The curious ones tinkering in the shadows. Pull them into a working group. Give them a platform. Let them lead from the middle.
Training Unlocks Access
Implement a plan for entire teams. Gamify it. Make progress visible. Leaderboards. Badges. Everyone can see who's completed AI certification and who hasn't.
Train the Gatekeepers
The EAs. The chiefs of staff. The people your execs actually trust. Enable them first. Let them show leadership what's possible over coffee.
The Journey You're About to Take
So what happens when you actually start? Not the pitch deck version. The real version.
First, It Will Be Uncomfortable
When you automate the process work, some of your people will freeze. The ones who built their career on execution will suddenly have nothing to execute. They'll sit in meetings wondering what their job is now.
This is normal. Expect it.
Strategic thinking becomes the new currency. The ability to solve problems AI can't see. The judgment calls. The human bit.
You'll need to redeploy, not just retrain. The new roles aren't "AI specialist". They're "problem finder + AI director." Promote those who get it first.
Then, Momentum Builds
Your team stops asking for permission and starts showing what they built. Slack channels full of wins. Champions teaching others. The curious ones who were tinkering in the shadows are now leading from the middle.
"You did that in three hours?"
"How long would it have taken before?"
"Three days!"
"Can we roll that out for everyone?"
"Can you build on it? Can we take it to clients?"
That staff member gets recognised. They feel empowered. They have purpose.
That's the shift. From hiding time to sharing wins.
AI Launch Roadmap: Four Weeks to Momentum
This isn't a twelve-month transformation plan. It's four weeks to get moving. Board approval. Exec alignment. Pilots running. Wins on the board.
Get the Board to Yes
Your only job this week is getting approval. One paper. One meeting. Forty-five minutes.
- Write a five-page board paper. Not fifty pages. Five.
- Pick three use cases with realistic time savings you can prove in 90 days
- Propose 10-20% of tech spend (0.5-1% of revenue) for pilots
- Address data privacy, ethics, vendor lock-in upfront
- Propose an AI steering committee with you as chair. Not the CIO. You.
Lock In Your Executive Team
Ninety minutes with your execs. No longer. Three outcomes.
- Run a data audit — can three teams actually access clean data?
- Identify 2-3 champions per function (already using AI)
- Shortlist three vendors max: ChatGPT Enterprise, Claude, one internal option
- Each exec leaves owning one pilot with clear deadlines
Clear the Path
Two workstreams running in parallel.
- Legal & Risk: Get vendor contracts signed. Data sovereignty clauses mandatory for NZ businesses.
- Communications: Draft internal message. Framing matters: "AI won't replace you. Hiding from it will."
- Announce shadow AI amnesty — companies that do this see 3x more shadow AI reported
- Have EAs run 30-minute AI training sessions for each direct team
Launch and Lead
This is where most companies stall. They plan forever and never start. You're not going to do that.
- Brief direct reports. 30 minutes. "The board approved $X. Each of you owns a pilot."
- Activate champions: "I need one win from you by Friday. Then teach your team."
- Use it yourself. Block 90 minutes. Real work. Not a demo. Your team is watching.
Budget Benchmarks
What leading companies are investing:
| Company Size | Pilot Budget | Scale Budget |
|---|---|---|
| $50M Revenue | $250K - $500K | $500K - $1.5M |
| $100M Revenue | $500K - $1M | $1M - $3M |
| $250M Revenue | $1.25M - $2.5M | $2.5M - $7.5M |
Leo Garcia-Curtis
Founder & Head of Operations, WABOOM AI
16 months in rooms with CEOs who've cracked AI transformation