The foundations of AI readiness and what needs to happen

I run a marketing business, which might seem like an odd background for writing about AI readiness. But I spend a lot of time inside technology companies and alongside the people trying to modernise them, and what I keep seeing isn’t a technology problem. It’s a problem with the foundations. 

As part of the Mahi Tahi collective, I work alongside specialists in data and BI, cloud infrastructure, cybersecurity, process improvement, and automation. Across those disciplines, the pattern is consistent: organisations are being pushed toward AI before the conditions for it to work are in place. And nobody in the vendor ecosystem has a strong incentive to tell them that.

So this is my attempt to say it plainly.

Pressure to implement AI, but the foundations are not ready

Everyone is being told to have an AI strategy. Boards are asking for one. CEOs are promising one. Vendors are selling one. And underneath all of that noise, most technology and operations leaders are quietly grappling with the same uncomfortable truth: the foundations aren’t there yet.

That’s not a criticism. It’s just where most organisations are right now.

AI adoption is being driven from the top down, which means the pressure to act is real, but the conditions for success are often missing. And the gap between “we’re doing AI” and “AI is actually working for us” comes down to a handful of things that don’t get enough attention in the vendor conversation.

Here are three challenges to explore:

• The data problem

AI doesn’t create insight from bad data. It amplifies it. Organisations with fragmented, inconsistent, or ungoverned data don’t get smarter when they adopt AI tools. They get faster at producing unreliable outputs. The unglamorous work of getting data structured, governed, and analytically sound isn’t a precondition to AI in some theoretical sense. It’s the difference between an AI investment that pays off and one that quietly disappoints.

• The infrastructure problem

Where does your data go when it enters an AI tool? For many New Zealand organisations using offshore platforms, the honest answer is: we’re not entirely sure. US-owned hyperscale providers have acknowledged they cannot guarantee protection from US government access to data stored on their infrastructure. In a world where AI tools are routing sensitive business data through offshore environments, that’s a risk most boards aren’t fully across. It’s one of the reasons NZ-owned sovereign cloud providers like TEAM Cloud are having very different conversations with clients than they were two years ago. Sovereign infrastructure isn’t a niche IT preference anymore. It’s a governance question.

• The process problem

AI initiatives don’t fail because the technology doesn’t work. They fail because the processes feeding them are inconsistent, the governance isn’t in place, and the organisation isn’t set up to change how it works. Good change management isn’t a soft add-on to a transformation programme. It’s what determines whether the investment sticks.

What good looks like

Organisations that get real value from AI tend to share a few things. Their data is clean and governed. Their infrastructure is secure and, increasingly, sovereign. Their processes are streamlined and consistent enough to automate. And their teams understand why change is happening, not just what is changing.

None of this requires a multi-year programme before AI can enter the picture. But it does require someone to ask the right questions in the right order before the technology conversation gets too far ahead of the foundations.

That’s the work most organisations haven’t done yet. And it’s the work that determines whether AI delivers on its promise or becomes another expensive lesson.

If you’re not sure where your organisation stands, it’s worth finding out before the board asks.

Mahi Tahi is a collective of six New Zealand technology and business services companies helping organisations build the foundations that make AI work.