The Return on Courage

I've spent much of 2026 standing in front of rooms full of people talking about artificial intelligence. Boards and executive teams. Conference halls. Schools and universities. Government departments. Nonprofits trying to stretch every dollar a little further. The audiences change, the sectors change and the questions are phrased differently, but by the middle of the year I had stopped believing that the hardest part of this moment is the technology.

The technology is extraordinary, and it is moving faster than most of us can comfortably absorb. But what AI has exposed, more than anything, is a leadership vacuum. In room after room, people are waiting for someone to tell them what this is for. Not how it works, or which tools to buy, or which policies to sign. What it is for, and who it is for. Too often, nobody answers.

This piece is my attempt to say what I have come to believe about that vacuum, and what I think it asks of leaders.

A room in Toronto

One of the most important rooms I stood in this year was in Toronto, at the YMCA World Council. A significant part of that audience was young people, and I wanted to speak to them directly rather than about them.

Many young people reject AI, and I understand why. For the past few years they have been told, loudly and repeatedly, that their education is obsolete, that the jobs they were preparing for will disappear and that the creative work they love can now be produced by a machine in seconds. Underneath all of that sits a quieter and more damaging message: that none of this is up to them. The future is arriving, it was designed somewhere else, and their role is to adapt to it.

If that is the story you have been told, rejecting AI is not ignorance. It is a reasonable act of self-protection. When someone tells you that you have no agency, refusing to participate is one of the few forms of agency left.

So I didn't try to talk them out of their scepticism. Instead I suggested that they have three options.

• — Reject AI. Step away from it, refuse to use it and hope the world that values what you value holds its shape.
• — Use AI solely for efficiency. Get your assignments done faster, produce more with less effort and treat it as a shortcut through the work.
• — Embrace AI to expand what you can learn, how well you understand yourself, what you can create and what you believe is humanly possible.

The first option is understandable, but it hands the shaping of the technology to everyone else. The second is where most organisations currently are, and it is a small ambition dressed up as progress. The third is harder, less certain and far more interesting. It treats AI as a way to become more capable and more yourself, not as a way to do less of the work that made you who you are.

Then I said the line I most wanted them to take home:

Leave optimising the past to the old guys. Go and build the future. Because frankly, we need you to.

I meant it. Not as a slogan, but as an honest assessment. The generation that grew up with this technology is better placed than mine to imagine what it could be for. What they lack is not ability. It is permission, and a story in which they are the authors rather than the casualties.

The story most organisations are telling

Back in boardrooms and executive offsites, the dominant AI story sounds very different. It is a story about productivity, automation, efficiency and cost. How many hours can we save? Which processes can we automate? How much can we take out of the cost base? Business cases are written in that language, and progress is measured against it.

None of those questions are wrong. Organisations should be efficient, and wasted effort is not a virtue. But listen to the story from the other side of the table. When employees hear their leaders talk about AI almost entirely in terms of efficiency, they hear one thing very clearly: you are the inefficiency. You are the cost we are trying to remove.

Then leaders are surprised by resistance. They describe it as change fatigue, or a skills gap, or a cultural problem to be managed with communications and training. Usually it is none of those. It is people responding rationally to a story in which they have no future. You cannot ask people to wholeheartedly adopt a technology they have been told, in effect, is being deployed to replace them. The resistance is not a failure of the workforce. It is a predictable result of the story leadership chose to tell.

Managing money is not the same as leading

I think the efficiency story has deep roots. For decades, many organisations have operated under intense short-term economic pressure: quarterly reporting, shareholder expectations, constrained budgets and the constant demand to do more with less. Over time, a quiet substitution has taken place. Financial management has come to stand in for leadership.

I want to be careful here, because financial discipline matters. Leaders who cannot manage money put their organisations, and the people in them, at real risk. Good stewardship of resources is a responsibility, not a lack of imagination. I have worked with many finance leaders who are among the most strategic thinkers in their organisations.

But financial management and visionary leadership are different things. Financial management asks how we can make the existing organisation perform better. Visionary leadership asks what the organisation should become, and why that future is worth the effort of getting there. The first protects the present. The second takes responsibility for the future. Both are necessary. The problem comes when the first is mistaken for the whole job, and AI is handed to people whose only brief is to optimise what already exists.

Under those conditions, AI becomes an extraordinarily powerful tool for doing the past more cheaply. That is a remarkable waste of a technology that could help us do things we have never been able to do before.

The future is a decision

Much of the public conversation about AI treats the future as something that will simply happen to us. The technology advances, and society adjusts. Jobs disappear or appear, institutions adapt or fail, and the outcome is described as if it were a weather forecast.

I don't accept that. The future of AI is not technologically predetermined. It is the consequence of human decisions. Decisions about how work is designed and who gets to do the interesting parts of it. Decisions about what and how we teach. Decisions about governance, about what we permit and what we prevent. Decisions about where investment goes. And, most importantly, decisions about who benefits.

Every one of those decisions is being made right now, in boardrooms, cabinet rooms, classrooms and planning sessions. Many of them are being made by default, because nobody thought they were a choice. A leader who frames AI purely as a cost-reduction tool has made a decision about the future, whether they recognise it or not. So has a leader who frames it as a way to expand what their people and their organisation can achieve.

This is why I keep saying that AI has exposed a leadership vacuum. Where nobody articulates a future worth building, the default future fills the space, and the default is usually a narrower version of the present.

The Return on Courage

This brings me to an idea I have been testing in conversations throughout the year. I call it the Return on Courage.

I want to be clear about what it is and what it is not. The Return on Courage is a hypothesis. It is not a proven financial metric, and I am not claiming to have measured it. It is a way of naming a pattern I believe is real and worth taking seriously: that the organisations most likely to gain from AI over time will be led by people with the courage to aim higher than efficiency.

In practice, I think that courage looks like a few things.

• — Using AI to expand human potential, giving people more capability, more judgement and more interesting work rather than simply less of it.
• — Using AI to extend the organisation's mission, asking what it could now do for customers, communities or the people it serves that was previously out of reach.
• — Enabling genuine experimentation, with guardrails that make it safe for people close to the work to try things, learn and share what they find.
• — Prioritising learning velocity, the speed at which an organisation can notice change, make sense of it, act and improve.
• — Being willing to change the operating model itself, rather than bolting new technology onto structures designed for a different era.

Each of those requires leaders to accept uncertainty they cannot fully control. Experiments fail. Some new ideas will not work. Redesigning how an organisation operates is slower and messier than automating a process. On a short enough timeframe, the efficiency story will almost always look safer.

My hypothesis is that over a longer timeframe it is not safer at all. An organisation that only optimises becomes very good at a version of itself that may no longer be needed. An organisation that learns, experiments and reimagines its purpose builds capability that compounds. If that is right, courage is not the opposite of prudence. It is what prudence looks like when the ground is moving.

I may be wrong about parts of this, and I expect the idea to sharpen as more organisations try it. But I would rather leaders tested a hypothesis about building a better future than accepted, without question, a story about shrinking the present.

Courage is not naivety

Courage in this sense is not optimism with the risks edited out. The concerns people raise about AI are serious, and leaders who dismiss them lose the trust they will need to lead through this change.

Some people will be displaced, and in some roles the disruption will be significant. The benefits could flow very unevenly, widening existing inequality between organisations, communities and countries. The technology raises real questions about privacy and about how personal information is collected and used. And there is a genuine risk that power, both economic and informational, becomes concentrated in fewer hands.

None of that is an argument against ambition. It is an argument for taking responsibility. Governance matters enormously, but its purpose should be clear. Good governance should not exist simply to slow everything down or to protect leaders from blame. It should protect the future we choose: the people we want to include, the values we refuse to trade away and the outcomes we are trying to create. Governance designed around a clear purpose can enable bold action. Governance designed around fear tends to produce paralysis, which carries its own risks.

The leaders I most admire hold both of these at once. They are honest about the harms, and they refuse to let the fear of those harms become an excuse for having no vision at all.

The question I'm leaving with you

When I think back across the rooms I've stood in this year, the most hopeful moments came when people stopped asking what AI would do to them and started asking what they could do with it. That shift is not technical. It is a shift in agency, and leaders have more influence over it than anyone.

The young people in Toronto deserve leaders who believe they can build the future rather than simply survive it. So do the employees in every organisation currently being told that AI is mainly about cost. So do the communities and customers who will live with the consequences of the decisions being made now.

So I'll finish with the question I have been asking leaders all year, and the one I keep asking myself.

Are you simply preparing your organisation for the future you think is coming, or are you prepared to take responsibility for building the future you actually want?