To AI or not to AI, that's the question?

Everyone is focused on finding new ways to use AI.
I'm more interested in where we choose not to use it.
However, many organizations are avoiding this conversation.
And I was wondering why. Let me share my reflections. 😉
AI is often presented as a technology decision. It isn't.
It's a risk decision.
Every AI capability we introduce also introduces uncertainty.
Unlike traditional software, AI doesn't simply execute predefined rules. It produces probabilities. That changes the conversation from "Can it do this?" to "What happens when it gets this wrong?"
Here is where many AI discussions fall short.
The business case is usually measured in productivity gains and cost savings.
The hidden costs rarely make it into the slide deck.
Governance
Oversight
Validation
Compliance
Accountability
Trust
These aren't implementation details. They are part of the investment and need to be considered as such.
Not every process benefits from AI simply because AI is capable of performing it.
When the cost of an error is low, automation makes sense.
When creativity is needed, AI can accelerate the first draft.
But when decisions affect stakeholders (=people), or involve ethical judgment, the equation changes completely. In those moments, AI should support human thinking and not replace it.
We need to understand AI boundaries.
In my experience, knowing where to use AI is becoming a technical competency.
Knowing where not to use it is becoming a leadership competency.
What are your thoughts?