AI conversations are happening in most organisations right now, and if you work in data governance, you're probably right in the middle of it.

AI is everywhere. Executives are excited, and rightly so. The possibilities are genuinely exciting: new services, reduced costs, greater efficiencies, things that simply weren't possible before. It's shiny, it's fast-moving, and there's enormous pressure to get on with it.

And then there's you. The data governance professional. Quietly clearing your throat and saying, "But wait..."

I get it. I really do. You’re not wrong to have concerns, but the way we raise those concerns matters enormously. If we're not careful, we become the naysayers. The people who slow things down. The ones standing in the way of progress with a clipboard and a risk register.

That's not who we are. And it's not who we need to be.

Stop Talking About Data Governance

I know that sounds counterintuitive, but hear me out.

Nobody (and I mean nobody) wakes up wanting to do data governance for the sake of data governance. What they want is for their AI initiative to be a success. They want the business outcomes that AI promises to deliver. So if we walk into a conversation leading with "we need to do data governance first," we've already lost the room.

Instead, start with the AI initiative itself. What is your organisation actually trying to achieve? What problem is this first AI project solving? Who are the stakeholders driving it? Go and talk to them, not to slow things down, but to understand what they're building.

Then work backwards. What data does that AI need to function? Is that data good enough? Do you know where it lives? Is it documented? Are there known quality issues? Could it be coming from multiple sources that need to be pulled together?

That's your story. Not "we need data governance." But "here's why your AI project won't deliver what you're hoping for unless we address these specific data issues."

Make It Specific. Make It About Outcomes.

The more specific you can be, the more persuasive you'll be. Vague warnings about data quality don't land. But "this AI initiative is designed to improve customer response times, and the customer data it will rely on has known duplication issues that will undermine the results" — that lands.

Tie every data governance conversation directly to the outcome the AI is supposed to deliver. If the data isn't fit for purpose, the AI won't achieve what the business is counting on. Say that clearly, calmly, and with evidence where you can.

Be Pragmatic, Not Perfect

Here's the other trap to avoid: if you do get the green light to work on data governance alongside the AI project, don't immediately demand that everything be perfect before anything moves forward.

That's a fast track to being sidelined.

Instead, think in phases. What's the minimum we need to get right for the testing phase? What can we improve incrementally as the project progresses? How do we raise the bar over time rather than trying to boil the ocean before day one?

Pragmatism isn't compromising your standards. It's recognising that progress beats paralysis, and that building trust with your AI colleagues is how you earn a permanent seat at the table.

Be the Enabler, Not the Obstacle

The shift I'd encourage every data governance professional to make is this: stop positioning yourself as the person who might block the AI project, and start positioning yourself as the person who will help it succeed.

Because that's the truth. Good data governance isn't the enemy of AI. It's what gives AI the foundation it needs to actually deliver. When the data is right, the AI works. When the AI works, the business gets its outcomes. And you were part of making that happen.

That's a very different conversation from "don't do AI without data governance."

Want to learn more about AI and Data Governance? Book a call with me below!

Book a Call