Lead Your AI. Don't Let It Lead You.

Many of us are using AI at work now — not because we're AI fans, but because it's become part of our job.

The scope of our work as leaders has also expanded. Leadership can no longer just be about leading self, teams, or organisational change.

We now must learn how to lead our AI too.

I've noticed many leaders have started to abdicate their leadership to AI. If you think AI is smarter, better and faster than us humans, you’d settle for being just a ‘human in the loop’ — babysitting and checking AI's work rather than directing it.

Without clear leadership, ‘Human in the Loop’ just creates human rubber stamps.

AI thinks, AI drives, the human just complacently goes along for the ride and becomes the convenient bearer of full responsibility should anything go wrong.

That’s not how any of us want to work.

I think it's better to be Human in the Lead. That means teaching people to drive strategy, intent and judgment — working with AI as diagnostic tool, not decision-maker.

Think of AI as a scarily intelligent, highly impressive intern you get to work with.

They always have a quick, articulate response for you. They are tireless workers. They’ve read way more than anyone else in the room.

But they’ve also inhaled a whole lot of questionable, inaccurate information. And they can’t always tell the difference. They also like sounding confident even when they’re unsure or wrong. Worse, they have a strange sycophantic streak: they rather please you than push back when they see you making a mistake.

Would you let this intern lead a project, create assets for your organisation or make decisions unsupervised?

Of course not. You'd rein them in. Clarify boundaries. Check their work. Remind your team to never mistake their fluency for wisdom.

For us to be more effective Humans in the Lead, we need 5 ABCDE elements:

  1. Accountability — to drive the strategy, intent and judgment behind AI's output

  2. Bandwidth — enough time and space to actually lead AI well

  3. Competency — language and metacognitive skills, plus understanding of how AI works

  4. Discernment — the moral wisdom to make ethical calls

  5. Expertise — the specific domain knowledge the task needs

Of the five, Bandwidth is the most overlooked.

Even leaders with accountability, competency, discernment and expertise can still let AI slop pass. Like AI itself, our judgment gets impaired when we work over too long a stretch of time and get increasingly overwhelmed and unfocused.

Helping our people develop these five elements is leadership work that won't go away.

That's why I developed Purpose Before Prompt.

It's a compilation of what I'm learning about how leaders can work better with AI.

I want to work on your Competency: sharpening how you talk to it, think with it, and understand how AI works as a technology and a business.

I also want to develop your Discernment: recognising the moral tensions you face as you lead your team through AI adoption and considering the ethical choices you must make.

I’ll help you put that into a practical, DIY ethical guardrail you can take away and adapt for your own AI. It’s something I've built for myself and am adapting for my team.

If this interests you,

Register for Purpose Before Prompt here (13 Jan 2027):

I would love to see you at training,

Shiao

Shiao-Yin Kuik

Executive Director of Common Ground specialising in culture change, leadership development, and social impact strategy for mission-driven organisations. Her work helps leaders bridge the gap between what they value and how they actually work.

Currently in Year 7 of a 9-year government-commissioned civic engagement project, her consulting draws on two decades building, closing, and rebuilding organisations with integrity.

Next
Next

Is your organisation suffering from a conspiracy of stupidity?