I have sat in enough boardrooms and senior leadership meetings to know what good decision-making looks like, and I have sat in enough of them to know how rare it actually is. Making a genuine decision, by which I mean committing to a course of action under conditions of real uncertainty, accepting the risk that the decision might be wrong, and being prepared to own the outcome either way, is one of the hardest things a professional is ever asked to do. It is also one of the things that is most clearly, and perhaps permanently, human.
AI can do many things that used to require significant human effort. It can model scenarios, surface risks, synthesise data, generate recommendations, and do all of this faster than any analyst or advisory team ever has. The quality of information available to leaders making important decisions has improved dramatically as a result. But here is what has not changed: someone still has to decide. Someone still has to weigh the factors that no model fully captures, take a position that commits the organisation to a direction, and carry the responsibility for what follows. That person is irreplaceable not because of what they know, but because of what they are willing to do.
This is the seventh and final article in the August section of this series. I want to use it to make an argument that might seem counterintuitive: that AI, by improving the information environment in which decisions are made, is actually raising the stakes for the quality of human judgment rather than reducing it. The decision is not getting easier. It is getting more visible, more consequential, and more demanding of genuine leadership.
This Series: 10 Skills AI Cannot Replace
July 2026
August 2026
September 2026
What AI Has Changed About the Decision Environment
Let me start with what AI has genuinely improved. The information available to leaders making important decisions is significantly richer than it was a decade ago. Scenario modelling that used to take days can now be done in hours. Risk analysis that required a team of specialists can now be produced rapidly by AI systems that have access to far more data than any individual analyst could process. The ability to rapidly surface precedents, model alternative outcomes, and identify risks that might have been missed in a manual process is real and growing.
This improvement in the information environment is valuable. It means that leaders who use AI well are operating with a better understanding of the landscape before they decide. But it has also created a new challenge that I see frequently in the organisations I work with: the temptation to keep seeking more information rather than making the decision. When AI can always generate one more analysis, one more scenario model, one more risk assessment, the moment of commitment keeps getting deferred. The decision-maker experiences this as being more rigorous. What is actually happening is that the difficulty of deciding, which is not about information but about the willingness to commit and be accountable, is being avoided under the cover of thoroughness.
"AI increases the speed of the operational environment. This means that as AI spreads, decision quality matters more, not less. The leader who can make good decisions quickly, in conditions of genuine uncertainty, and own the outcomes, becomes disproportionately valuable precisely because the cost of poor or delayed decisions is rising."
The Five Leadership Capabilities That AI Cannot Replace
Judgment: knowing what matters most when everything matters
Judgment is the ability to weigh factors that cannot all be fully quantified and arrive at a conclusion that is better than what any model would produce, because it integrates information that no model has access to: the specific context, the relationships involved, the values of the organisation, the things that were said off the record in last week's conversation, and the intuition built from years of operating in similar environments. AI models optimise for what they can measure. Judgment operates across the full reality of the situation, including the parts that cannot be measured. This is why experienced leaders with genuinely good judgment remain, in my observation, the scarcest resource in most organisations.
Risk assessment: carrying the uncertainty that models cannot resolve
AI can model risk. It cannot carry it. Risk assessment as a leadership capability is not just about identifying what could go wrong. It is about being willing to make a decision in the knowledge that something could go wrong, having weighed the probability and consequence of different outcomes, and accepting responsibility for the result. The human element in risk assessment is not the analysis. It is the willingness to commit despite the uncertainty that the analysis cannot fully resolve. That willingness is a character capacity, built over time through the experience of making decisions, seeing outcomes, and developing the emotional resilience to try again when things do not go as planned.
Prioritisation: choosing what not to do
In an environment where AI can generate unlimited options and analysis, the leadership skill of prioritisation becomes more important, not less. Prioritisation is not about identifying what is good. It is about knowing what is most important given the specific constraints of this organisation at this moment, and having the conviction to say no to everything else, including things that are genuinely good. This is an act of leadership that requires a clear values hierarchy, an understanding of what the organisation is actually trying to accomplish, and the courage to disappoint people who wanted resources allocated differently. AI cannot provide any of these things.
Ethical decision-making: knowing what the right thing is when the right thing is costly
The ethical dimension of leadership decisions is one that AI is particularly ill-equipped to handle. Not because AI cannot identify ethical frameworks or generate analysis of the ethical implications of a choice, but because ethical decisions in real organisations involve the willingness to absorb a personal cost for the sake of doing what is right. The leader who tells the board something they do not want to hear because it is true. The executive who declines a profitable course of action because of its effect on employees or communities. The manager who gives honest feedback to a high performer because their team needs to hear it even though the conversation will be uncomfortable. These are acts of ethical leadership that require moral courage, not information processing.
Accountability: owning the outcome regardless of what the model said
Accountability is the most fundamental leadership capability, and it is the one that AI cannot participate in at all. When a decision goes wrong, there needs to be a human being who is responsible: who acknowledges what happened, who understands why, who protects the team members who executed faithfully, and who develops the organisation's capacity to do better. This is not a capability that can be delegated to a system, however sophisticated. Organisations run on accountability. Without it, nothing else works. And accountability is entirely a human act.
A Real Example from the Corporate World
A leadership team faces a restructuring decision with significant human consequences
The data is clear. The AI-powered analysis has modelled twelve scenarios, with projected financial outcomes, risk assessments, and implementation timelines for each. The recommendation is unambiguous. A specific restructuring will improve the organisation's financial position significantly and position it better for the next three years.
But the data does not capture everything the leadership team needs to weigh. It does not capture the long-term culture consequences of the restructuring approach being recommended. It does not capture what will happen to the team members who will be most affected, or what the impact will be on the organisation's ability to recruit in a specific market going forward. It does not capture the fact that one of the options, which scores lower on the financial model, is actually more aligned with the values the organisation has publicly committed to.
A thorough, well-modelled analysis of twelve restructuring scenarios with projected financial outcomes, risk assessments, and implementation recommendations based on available data.
The judgment about which factors the model had not adequately weighted. The ethical decision about what the organisation's values actually required. The courage to choose a less financially optimal option for the right reasons. And the accountability for the outcome either way.
The AI analysis was genuinely useful and improved the quality of the conversation. But the decision was not made by the model. It was made by human beings who weighed what the model could not weigh, and who were prepared to stand behind what they chose.
How to Develop Leadership Judgment Deliberately
Judgment develops through the cycle of deciding, observing the outcome, and honestly examining the gap between what you expected and what happened. Professionals who avoid decisions to protect themselves from being wrong are also preventing the development of the judgment that comes from experiencing outcomes. Make decisions, including ones you are not certain about, and build the discipline to review them without defensiveness.
The leaders who make the best decisions under pressure are almost always the ones who have done the work of clarifying their values before they face a situation that requires them. Knowing what you will not compromise on, what the organisation's non-negotiables are, and where the ethical boundaries sit, before a specific decision puts those questions under pressure, is what makes ethical leadership possible in real time. This is reflective work, and it requires genuine honesty rather than abstract principles.
One of the most important things a leader can learn is the difference between genuine uncertainty, where more information would materially change the decision, and the discomfort of commitment, where the information is sufficient but the psychological act of deciding is what feels difficult. Building the capacity to recognise which situation you are in, and to commit when the information is sufficient, is a judgment skill that develops through deliberate practice.
Accountability is not a personal quality that exists in isolation. It is a culture that leaders build through their own behaviour. When you own a mistake clearly, without deflection, and focus on what the team can learn from it, you build an environment where others feel safe to do the same. This culture, where accountability is the norm rather than the exception, is one of the most valuable things a leader can create, and it is created entirely through human behaviour.
The most valuable thing coaching offers a leader is the experience of having their thinking challenged by someone who is not invested in the outcome. A coach who asks hard questions about your assumptions, your values, your blindspots, and your willingness to be accountable before a decision is made can materially improve the quality of that decision. In 28 years of corporate life and a decade of coaching, the leaders I have seen make the best decisions consistently are the ones who have invested in developing their judgment through exactly this kind of structured challenge.
"Executives are paid for judgment, risk assessment, prioritisation, ethical decisions, and accountability. As AI increases the speed and volume of the decisions that organisations need to make, and raises the quality of the information available to make them, the demand for leaders who can genuinely exercise these capabilities is not declining. It is growing. Leadership is becoming more valuable, not less."
The August Series: What We Have Covered
This article closes the August portion of this series. Over the past four articles, we have looked at creativity and original thinking, AI collaboration skills, adaptability and learning speed, and leadership and decision-making. Each of these capabilities shares a common quality: they are rooted in specifically human capacities, they become more valuable as AI takes on more of the execution and information-processing work that previously required human effort, and they are all developable with the right investment and the right support.
The September articles will look at the final three skills in the series: Domain Expertise Combined with AI, Systems Thinking, and Ethics, Trust, and Governance. These three skills are, in my view, the ones that will define the most senior and most consequential professional roles over the next decade. I hope you will continue with us.
Leadership Judgment Is the Investment That Compounds
Of all the capabilities in this series, leadership judgment is the one that compounds most powerfully over time. Every decision you make well, every accountability you carry with integrity, every value you hold when holding it is costly, these build something that no AI system can build for you. Coaching is the fastest and most direct way to develop it.
Book a Free Conversation