The AI Audit: What Happens When You Turn the Lens on Yourself

After the HYROX race, I sat in the car park for a while before driving home. Not because I was too exhausted to move, though I was, but because something had shifted.

I'd spent the last two years telling boardrooms across Asia-Pacific that AI would transform how organisations operate. That leaders who didn't build AI into their workflows were already behind. That clarity came from having the right intelligence infrastructure, not just the right instincts.

And then I looked at my own life. My actual life. Not the keynote version of it.

I wasn't using AI as infrastructure. I was using it as a search engine with better manners. I'd ask it to draft an email. Summarise an article. Clean up a paragraph. Occasionally, when I was feeling adventurous, I'd ask it a strategic question, and then largely ignore what it said.

So I decided to audit myself.

Not my organisation. Not my clients. Me. My finances, my health, my decision-making, my parenting, my professional practice, everything. I wanted to know where AI was genuinely helping me think better and where I was just performing digital fluency without actually doing the work.

The results were uncomfortable. And that's exactly why I'm writing about them.

The gap between talking about AI and living with it

There's a particular kind of hypocrisy that's easy to slide into in my industry. You become fluent in the language of transformation. You build frameworks. You earn the right to stand on stages. And somewhere in that process, you stop being the student.

I had stopped being the student.

My AI usage had calcified into a set of comfortable, low-stakes habits. Ask, receive, move on. Never linger. Never push back on the answer. Never say "go deeper," "challenge that," or "what am I missing here?"

The audit started simply: I went back through three weeks of my AI conversation history and categorised every single prompt.

The breakdown was illuminating. Roughly 60% of my prompts were administrative, drafting, formatting, and summarising. About 25% were informational, essentially using AI as a faster search engine. And maybe 15% were what I'd actually call thinking, using AI to challenge an assumption, pressure-test a decision, or see a situation from a different angle.

Fifteen per cent. For someone who makes a living telling leaders that AI is a thinking partner, not a typing assistant, that number was hard to sit with.

Tool vs thinking partner: the distinction that changes everything

When I talk to executives about AI adoption, I use a simple frame: there's a difference between using AI as a tool and using AI as a thinking partner. A tool does what you ask. A thinking partner challenges what you're asking for and why.

Most organisations are firmly in tool territory. They automate the mundane. They speed up the routine. They get more output from the same input. That's genuinely valuable. But it's not a transformation.

Transformation happens when the AI makes you think differently. When it reflects something back at you that you didn't see before. When it asks the question you were avoiding.

My audit told me I was in tool territory. And I decided to change that, domain by domain, starting with the areas of my life where I had the most resistance.

Money first. Because money, as it turns out, is where I was hiding the most.

Practical AI: Audit your own AI usage

Before the next post, try this. Go back through your last week or two of AI interactions, whatever tool you use, and categorise each prompt into three buckets:

• Administrative, drafting, formatting, organising, summarising

• Informational, research, fact-checking, learning about something

• Thinking, challenging assumptions, stress-testing decisions, gaining perspective

What's the split? Most people, when they're honest, find that the thinking category is surprisingly thin. That gap is exactly where the opportunity lives.