Imagine you're told that employees aren't engaging with a new organisational initiative.
There are plenty of obvious places to start. Improve the communications. Create clearer messages. Equip managers to talk about it. Produce FAQs. Develop a campaign to build awareness and understanding.
AI could help with all of those things.
But spend some time with employees and you might discover that awareness isn't the issue at all.
Perhaps people understand the initiative perfectly well but don't believe anything will actually change. Maybe previous initiatives have quietly disappeared. Perhaps their day-to-day experience contradicts what leaders are saying. Or they might simply have other, more immediate concerns competing for their attention.
Suddenly, better communications don't look like the obvious answer.
That's why the way we frame a problem matters so much.
One of the great advantages of AI is that it removes some of the friction involved in getting from problem to output.
But sometimes that friction was useful.
It gave us time to think. To question the brief. To speak to people. To notice that different groups were experiencing the same situation differently.
When generating a credible response becomes almost effortless, there's a risk that we skip too quickly over that part.
We ask AI to solve the problem we've been given rather than helping us explore whether it's the problem we should be solving.
For EX practitioners, that distinction matters.
Perhaps one of the most valuable ways we can use AI isn't simply to generate solutions, but to improve our thinking before we get there.
We can use it to challenge our assumptions. Generate alternative explanations. Identify whose perspective might be missing. Explore what we'd need to learn before making a decision. Look for patterns in qualitative feedback. Develop different hypotheses that we can then investigate with real people.
In other words, AI doesn't have to shorten the journey from brief to answer.
It can help us make the space between them more useful.
And that's where human-centred approaches become even more important.
Talk to the people affected. Observe what actually happens. Understand the context around the problem. Involve people in developing possibilities. Make ideas tangible and test them before investing heavily in a solution.
None of that needs to mean months of research or a huge design project. Sometimes a handful of conversations can completely change how you see the challenge.
As AI becomes part of everyday work, knowing how to generate a good answer will become less distinctive.
Knowing which questions are worth answering may become much more so.
That's where our judgement as EX, HR and internal communication practitioners matters.
We need to know when to move quickly and when to investigate. When an obvious solution really is obvious and when the brief needs challenging. When AI can accelerate the work and when we need to get closer to people's actual experience first.
So perhaps the question to keep close isn't simply:
What should we do?
It's:
What might we need to understand differently before we decide?
