AI Can Recommend Your Next Move. It Cannot Own It.
14 Oct 2026
AI can recommend your next move. It cannot answer for it.
AI can recommend your next move. It cannot answer for it.
It drafts the analysis. It models the scenarios. And it sounds convincing every single time.
That is exactly the moment a leader is needed. And it's not only me, saying that.
A piece on the World Economic Forum site last week put it in one sentence: the measure of an AI-era leader will be the decisions they have the courage to question and the consequences they are willing to own.
Full article: https://www.weforum.org/stories/artificial-intelligence/leadership-ai-governance/
Question. Own. That is your judgment.
PS. This is the message of my new keynote series, "Give away the work. Never the judgment." This autumn I am taking it to more than 10 organisations, as a keynote and as a workshop. If you would like it for your organisation, the details are here: https://www.thechangerepublic.com/speaker. Or contact us directly.
The numbers point the same way. In the WEF's previous Future of Jobs Report 2025, over 1,000 employers across 55 economies were asked which skills are core. Leadership and social influence rose 22 percentage points compared with 2023. The biggest rise of any skill.
So how do you keep it?
→ Ask who carries the risk if the answer is wrong.
→ Decide upfront which calls stay with you.
→ When you overrule the tool, say why.
When did you last overrule AI?
Want to think through what AI means for your own leadership? My Private AI Advisory is here.
What '22 percentage points' actually means in a coaching room
That WEF number, leadership and social influence up 22 percentage points as a core skill, is not an abstract trend when you sit with it in a coaching session. It shows up as a very specific discomfort: senior people who are excellent at the technical and analytical side of their job, and who have never had to build the muscle of overruling a confident, well-written recommendation.
AI makes this harder in a quiet way. A bad human recommendation usually has some tell: hesitation, a gap in the argument, a nervous "I think". AI output has none of that. It reads as finished and sure of itself whether it is right or wrong, which means the signal leaders used to rely on to know when to push back is gone.
The three questions in the post above, who carries the risk, which calls stay with you, why you overruled it, work because they do not depend on spotting a tell. They are a structure you apply before you read the recommendation, not a feeling you wait for after.
→ Build the habit before the stakes are high. Practise overruling AI on a low-risk decision so the muscle is there for a high-risk one.
→ Write down your "why" even when nobody asks. The habit of explaining a call to yourself is what keeps it sharp.
→ Separate speed from certainty. AI gives you speed. Only you can supply certainty.
I wrote more about building this kind of AI fluency in what senior leaders actually need from AI in 2026.
If you want a structured way to practise this kind of decision-making, I have a free AI Coach Tünde Light, a set of self-coaching prompts for exactly these moments. Leave your email here and I will send it over: https://www.thechangerepublic.com/free-resources