Back to @tgroenwals's timeline
tgroenwals shared this post · 5h ago
LinkedIn
Clare Kitching
Clare Kitching

AI governance is getting the attention.
Data governance is doing the work.

Right now, companies everywhere are rushing to scale AI.

But underneath the excitement is a quieter question:
Can we actually trust the data feeding these systems?

Because AI is only as good as the foundation beneath it.

If nobody knows:

  • where the data came from
  • who owns it
  • whether it’s accurate
  • whether it’s current
  • whether sensitive information is buried inside it
    ...then AI will scale this confusion.

I’m seeing more leaders recognise that “bad data means bad outputs.”
But many still struggle with what to do to address it.

That’s why I think data governance is becoming one of the most important and underrated skills in the AI era.

Good governance creates trusted data, responsible AI, better business decisions and confidence at scale.

And data governance is hard.
It requires a combination of ownership, relationships, accountability, standards and trust.

It’s hundreds of small decisions that quietly determine whether AI creates value or risk.

AI governance may dominate the conversation.
But data governance is the unsung hero behind every successful AI strategy.

♻️ Repost to shine a light on data governance.
🔔 Follow Clare Kitching for insights on unlocking value with data & AI.

Khwaja Naveed The 'treat governance as AI infrastructure' point deserves more attention in AI strategy discussions. Most organizations build AI capabilities on top of poorly governed data and then wonder why their models underperform or produce inconsistent outputs. You can't retroactively govern data after the model has been trained on it. The accountability structure, ownership, and quality standards need to be in place before the AI use case is designed, not after it's already been deployed.
Victoria Kurashina AI governance frameworks assume the data layer already has an owner. Most companies are writing that policy on top of data nobody has actually claimed, which is why it reads fine and still doesn't hold.