You're asking the wrong question about AI

Hi Reader,

This week, a company asked me to help them build an AI policy.

And they kind of expected me to ask them about the tools they use, the rules they apply, and how they can check whether people are actually following those rules.

All reasonable questions. But I didn't start with any of them.

The first three I asked were:

How do your people feel about AI?

How do you keep it from getting in the way of their judgment?

And how do you plan to work with it, day to day?

Here’s why. A policy belongs to one dimension of AI transformation — how you work with the technology. It’s a real and necessary part. But it’s one slice of one dimension. Build the policy without touching the other two, and you end up with a document, not a transformation.

AI transformation has three dimensions. How people feel about AI. How it changes the way they think. And how teams work and collaborate with it.

I wrote three articles in my Forbes Coaches Council series, each on one dimension. And here’s what each of them is about:

How your people feel about AI. Fear rarely shows up as fear. It looks like caution — “we need more information before we decide.” It looks like a team that signs off on the rollout and never quite uses it. Left unnamed, it makes a transformation too slow and too fast at the same time: racing in some corners, stuck in others, with no shared sense of why. I wrote about this in The AI Transformation Gap: Managing Fear and Resistance, the first article in my Forbes Coaches Council series.

How AI changes the way they think. This one shows up when AI works. The team gets faster, frees up time, and slowly stops holding the thinking that used to live in their heads: the feel for what a client needs, the instinct for what matters. It looks like progress until performance starts to slip. That’s cognitive dependency, and it’s the subject of the second article, The AI Transformation Gap: Preventing Cognitive Dependency.

How you work with it. This is where the policy lives, and it’s a bigger question than which tools and which rules. Who owns which decisions. What always gets checked. What never gets delegated. When AI hands something back to a human. A policy is part of the answer. The norms that make people use it are the rest. I unpacked this in the third article, The AI Transformation Gap: Building AI Governance.

You can write a clean policy, and if your people are uneasy, they won’t follow it. You can write strict rules, and if their judgment is thinning out, those rules are guarding a process nobody’s really thinking through anymore.

So if you’re building an AI policy right now, and you want to bring structure and clarity to how your team is using AI, make sure to include all three dimensions in your research and only write your policy when you have answers to all three questions.

See you next week.

Daria

P.S. Most of these conversations start on my podcast, Built by People Leaders, where HR and people leaders tell me what AI adoption actually looks like on the ground. Worth a listen if this is live for you right now.