26/02/2026
AI is having its moment. And to be fair, it is amazing at the stuff that slows people down - ideation, first drafts, automation, turning a messy thought into something usable. It’s basically a shortcut through the blank page.
But here’s what’s becoming obvious:
AI still isn’t a one-stop shop. Most “real work” is a loop ... AI does a chunk, a human makes a judgement call, another tool gets involved, context gets layered in, then someone sanity-checks it against reality.
That’s why the organisations getting value aren’t the ones with the most licences. They’re the ones with the strongest human-in-the-loop capability.
Because the biggest blockers aren’t necessarily technical. They’re likely to be human.
People aren’t just thinking “how do I use it?”
They’re thinking: Is this cheating? Am I allowed? What if it’s wrong? Will I look lazy? Who’s accountable if it goes sideways?
That’s where leadership comes in.
If leaders don’t activate the climate for AI (permission, clarity, role-modelling, and a grown-up approach to risk), people either avoid it entirely… or use it quietly and inconsistently, which is how quality problems creep in.
Thing to try: “The Human-in-the-Loop focus” (10 mins, in any team meeting)
Agree this out loud as a team:
“We can use AI for speed… but humans own the judgement.”
Then set three simple rules in plain English:
1. where AI is encouraged (ideation, drafting, summarising, admin work)
2. where humans must take the wheel (decisions, customer impact, risk, sign-off)
3. what “good use” looks like (say when you’ve used it, sanity-check, don’t blindly paste copy, etc)
It sounds small, but it does something massive... it removes the awkward uncertainty that stops adoption and it protects quality!
In other words: we don’t need humans out of the loop. We need humans in charge of the loop.
If you want AI to lift performance (not just output volume), stop obsessing over tools and start building the climate that makes smart adoption normal.