“Most IT managers are not resisting AI. They just don’t want to be blamed for a bad rollout.”

There’s a narrative around that IT teams are slowing AI adoption.

That they’re being cautious. Too risk-aware. Blockers!

In most organisations ,that isn’t what’s really happening.

What’s happening is that IT teams have been told: ‘We need to do something with AI’. But they don’t have a definition of what ‘good’ looks like.

‍

And then …

If AI is rolled out too slowly, they’re seen as resistant.
If it’s rolled out too quickly, and something goes wrong, they’re accountable.

So they sit in the middle of an impossible balance:

Move fast, but don’t break anything.
Enable innovation, but stay fully compliant.
Support the business, but control the risk.

Meanwhile, AI is already being used.

Not through official. channels, but quietly across the organisation.

People are experimenting with public tools.
Uploading data they probably shouldn’t upload.
Building workflows that no one else can see or govern.

And this is where the real risk sits. Not in the technology itself, but in the lack of visibility around how it’s being used.

And when something eventually does go wrong, everyone says: ‘Well, why didn’t IT have control of this?’

‍

That’s the tension many IT managers are dealing with at the moment.

How to adopt AI in a way that is visible, controlled and defensible.

What ‘better’ looks like isn’t about slowing things down.

It’s about creating enough structure to enable AI to be used confidently.

‍

When guardrails, visibility and defined ownership are in place, the dynamic changes.

‍

AI stops feeling like something that could create problems. And starts becoming something commercially valuable, delivering measurable returns on everything you invest in it.

‍