What happens when senior leaders use AI to audit their own role?
Summary: Most senior leaders are using AI to improve the work of their teams. Fewer have thought to turn it on themselves. This post shares a live exercise run by Spark AI co-founder Emma Wharton Love, where senior leaders used AI to audit their own roles — and left with a personal action plan
Date: 27.05.2026
Author: Asta Vallis
Spark AI is a strategy-led consultancy helping agency and brand teams move from fragmented experimentation to organization-wide capability. Our blog provides the strategic techniques, insights and industry discussions needed to navigate AI with confidence.
Are you using AI to improve everyone's work except your own?
Most people are using AI to do work for others. Writing briefs, researching competitors, planning campaigns. It's a natural starting point, but it misses something. The same tools that help teams work more efficiently can do the same for the people leading them. Most leaders simply haven't thought to turn AI on their own role yet.
Last month, Spark AI co-founder Emma Wharton Love tried it with her peer group. She hosts a monthly group for senior women in the industry, sessions that usually focus on AI workflows and shared experience. Emma ran a live five-minute exercise: use AI to audit your own role and daily output. Every person left with a personal AI action plan, and the reaction was good enough that we wanted to share it more widely.
Why is the AI role audit worth doing?
Senior leaders are often the last people in an organisation to examine their own workflows. There's always something more pressing, and the assumption tends to be that efficiency gains are something to find further down the team. Asking AI to look at how you're spending your time changes that quickly. It surfaces tasks where you're spending more time than you need to, areas where the quality of your output could be higher, and opportunities to build tools that work specifically for your role.
How do you run it and what do you do with the output?
Open whichever AI tool you use most regularly, whether that's ChatGPT, Claude, Gemini or Copilot. Start by asking this question: “based on what you know about me, how could I be using you better to support my work?”
If it already has history from previous conversations, that's a good foundation. But if you're starting fresh, tell it your job title, the type of organisation you work in, your main responsibilities and what a typical week looks like in practice. The more specific you are, the more useful the output will be.
From there, ask it these three questions:
First, “identify tasks I do regularly where I probably spending more time than necessary, and explain how AI could speed each one up.”
Second, “suggest areas where AI could help me produce higher quality or more detailed work within the time I already spend.”
Third, “recommend a personalised toolkit I could build (within your chosen tool), and a clear purpose for each one.”
If the output feels generic, push back. Ask the tool to interview you about your biggest bottlenecks before it responds. A good set of targeted questions will produce far more useful recommendations than a broad prompt alone.
Once you have the output, treat it as a starting point. Pick one or two ideas that feel immediately actionable and build from there. The goal is a personal AI action plan you can start using this week.
Frequently asked questions
Is this exercise only useful for senior leaders?
The exercise works at any level. It tends to be most valuable for senior leaders because their time is pulled in more directions, but anyone whose role involves a mix of strategic and operational work will find it worth doing.
Do you need to use AI regularly for this exercise to work?
Regular use helps because tools with more conversational history give more specific output. Starting from scratch still works, but you'll need to spend a few minutes giving the tool context about your role before it can give you something valuable in return.
What if the output feels too generic?
Tell the tool the response isn't specific enough and ask it to interview you about your biggest time pressures before it tries again. The more targeted its questions, the more useful the output will be.
Turn fragmented AI experimentation into organisation-wide AI capability – with impact, control and confidence. https://www.wearespark.ai/
