You're Not Behind. Almost No One Is Ahead.

“Most leaders are sure they are ahead, or sure they are behind. Almost none can actually tell.”

Sit across from enough business leaders and you meet the same two people.

One is confident. They are well down the track with AI. The tools are in, the team is using them, it is working.

The other is anxious. They have read that nearly nine in ten firms are on board, and they feel like the last one to the party.

They are usually both wrong, and for exactly the same reason. Neither can see what good actually looks like. So one reads the visible tools as progress, and the other reads the headlines as proof everyone else has cracked it. Same blind spot. Opposite conclusions.

Push a little on the confident one. Ask what has actually changed about how the work gets done, not which tools are switched on. Usually they cannot say. There is a pilot running somewhere. A few people have found their own tools. There was training back in autumn. But no single process they can point to that runs differently today than it did a year ago.

The number that starts the angst

You may have seen the number that feeds the anxious amongst us. Datacom's 2025 State of AI Index reported that 87% of New Zealand organisations now use AI, up from 66% the year before. Read that as a leader and it lands in the gut: everyone else is doing this.

Then look under the hood of the same report. Only 12% have rolled AI out across the business. Nearly half are still poking at pilots. Just 8% call it transformative to how they operate. So the real message isn't "87% have cracked it." It's "87% have switched something on, and almost none have changed how they work."

And keep in mind who produces these surveys. Many come from firms with a commercial interest in AI, so the headline numbers deserve a careful read. Even then, only 8% claim to have transformed anything, and that figure is self-reported. Treat it as the ceiling, not the floor.

The number can't even agree with itself

In the same year, MYOB's Business Monitor surveyed a broader base, the small and medium firms that make up most of the economy, and put adoption at 32%, not 87%. Same country. Same twelve months. A fifty-point gap.

Both numbers are "right." They just measure different things. Switch on one Copilot licence and you count as "using AI" in one survey and not the other. When the measure is that loose, the headline cannot tell you where you stand, and a leader who trusts it makes real decisions on a false reading.

Globally it is the same story. McKinsey's 2025 research found 88% of organisations use AI somewhere, but only around one in five had fundamentally redesigned at least some workflows. Redesigning the work is what correlates with getting value. Almost no one is doing it.

The comfort is the trap

Back to the confident one, because their position is the more dangerous of the two.

It feels like progress, because the tools are visible. The licences are bought, the logins exist. But tinkering without a plan is not progress. It is noise. Lots of motion, nothing you can point to. That is why it will not scale, and why, a year in, you start quietly asking whether the spend was worth it. The tools created activity, not outcomes, and now you cannot tell the difference.

The real ladder has three rungs.

- Access: the tools are available.

- Usage: people use them for small tasks.

- Changed work: a real process now runs differently because of AI.

Only the third one shows up in your results. Almost every business reporting "AI adoption" is sitting on the first two, and cannot see the rung they are missing.

So the board question has to change. Not "how many of our people use AI?", but "which of our workflows now run differently because of it?". That measures change. One tells you what you bought. The other tells you where you actually are, and it will tell you more than any licence report.

Why the gap is this wide

The reason is not the technology. It never was.

Buying a licence is easy, fast and measurable. Changing how people work is slow, invisible on a dashboard, and runs straight into the hardest thing in any business: people, culture and the willingness to do the job differently tomorrow than you did it today.

AI adoption is a people-change problem wearing a technology costume. The bottleneck was never the software. It is the human capacity to absorb change. So the easy half gets done and counted, and mistaken for the whole. That is why the numbers climb while almost nothing changes underneath them.

What this means for you

You are not behind. Almost no one is ahead. The race you thought was nearly over has barely started for practically everyone in it.

That should not make you comfortable. It should make you move.

Because real adoption is rare, it is a genuine edge, not a game of catch-up. Because the missing rung is capability - not spend, the barrier is lower than most leaders fear. Lower, not easy. It costs attention, change effort and leadership discipline, not a bigger budget. The window is open for one reason: most organisations are still performing adoption instead of changing the work.

The same two reactions live inside your own head about your own use. A 2026 Qualtrics survey found only 12% of New Zealand workers stick to the AI tools their employer provides, the lowest rate of any country surveyed. Everyone is privately using their own, and guessing whether that puts them ahead or behind. They cannot see their own rung either. Nobody handed any of us the map.

Start here

1. Place yourself honestly.

Not "do we use AI?" but "what work is genuinely being done differently because of it?" If the honest answer is "a few faster emails," you are on the bottom rung. That is fine. Knowing it is the whole game.

2. Change your metric.

Stop counting licences deployed. Start counting workflows changed. One redesigned workflow beats fifty dormant subscriptions.

3. Take one rung, deliberately.

Pick a single process that repeats every week: sales proposals, client onboarding, board reporting, meeting follow-up, recruitment screening. Map how it runs today. Rebuild it around AI with the people who do the work. Measure the cycle time, the quality, the rework. Prove it, then decide whether to expand. That is what almost no one is doing, which is exactly why it works.

"You are not late. You are early, along with nearly everyone else. The only question is who stops guessing and starts changing the work. "

Find out where your AI adoption really stands

Most businesses are either overestimating their progress or assuming they are further behind than they are.

Take the 5-Minute AI Readiness Snapshot

--

About the Author 

Mark Lucas is Co-founder at Generation AI, where we help New Zealand business leaders build real AI capability.

Information provided is general in nature and does not constitute legal advice