In April 2026, I co-chaired a session for EGN’s Executive Leadership peer group in Singapore. We were at Marina Bay Sands — a fitting venue for a conversation about transformation, surrounded as it is by a skyline that looks like someone decided the rules didn’t apply.
The topic: The New CXO. What does senior executive leadership look like in an era when AI is reshaping the work, the decisions, and the nature of organisational advantage?
It’s a question I’ve been circling for the past couple of years — in classrooms, in boardrooms, in coaching sessions. The EGN conversation crystallised some things I’d been thinking about, and complicated some things I thought I understood.
The Autonomy Paradox
The first theme that emerged — and it recurred throughout the session — was what I’d call the autonomy paradox.
AI systems are getting genuinely capable. They can synthesise information faster than any human, identify patterns across data sets that would take teams of analysts weeks to work through, and generate options for complex decisions in minutes. The temptation — and in some organisations, the reality — is to let the AI lead and have the human ratify.
The problem is that when things go wrong — and they do go wrong — human accountability is still the foundation of institutional trust. Customers, regulators, employees, and boards all ultimately hold people responsible. You cannot delegate accountability to an algorithm.
So the CXO faces a genuine tension: leverage AI’s capability fully without losing the oversight and judgment that defines executive responsibility. The executives in the room were navigating this in real time, without a clear playbook. Several described moments of genuine discomfort — situations where the AI recommendation was compelling, but something felt off, and they weren’t sure whether that feeling was wisdom or resistance.
Human-in-the-Lead, Not Human-in-the-Loop
One formulation that landed well in the session was the distinction between human-in-the-loop and human-in-the-lead.
Human-in-the-loop is the defensive posture — AI does the work, a human reviews and approves. It preserves legal accountability but doesn’t necessarily preserve good judgment. If the human reviewer is simply rubber-stamping outputs they can’t meaningfully evaluate, the loop provides the appearance of oversight without the substance.
Human-in-the-lead is different. It means the executive defines the problem, sets the criteria, evaluates the outputs, and owns the decision — genuinely, not ceremonially. The AI is a powerful tool in service of that judgment, not a replacement for it.
The shift requires something specific from leaders: they need to be better at asking the right questions than at knowing the right answers. That’s a different cognitive skill, and it’s not one that most executive development has historically focused on.
What Doesn’t Change
Here is the thing that struck me most in this conversation.
We spent probably a third of the session talking not about AI at all, but about the fundamentals of leadership that remain stubbornly, enduringly human.
Presence matters. When a leader walks into a room — physically or virtually — the quality of their attention signals to everyone present how much this matters. AI cannot replicate that. It can summarise the meeting notes; it cannot create the felt sense of being heard.
Trust matters. The executives talked about how trust is built, how fragile it is, how difficult it is to rebuild when broken. AI systems can optimise for efficiency and accuracy. Trust is built through consistency, follow-through, and the willingness to be honest when honesty is inconvenient. That is still entirely a human domain.
Authenticity matters. One of the most interesting observations in the room was that as AI-generated content becomes pervasive — in communications, in reports, in strategies — authentic human voice becomes more distinctive, not less. The leaders who will stand out are those who communicate with genuine perspective, who have something to say that reflects actual experience and actual judgment. The polished, hedged, committee-approved communication style that has dominated corporate communications looks increasingly inadequate alongside leaders who speak plainly and specifically.
The Question I Left With
After the session, I found myself reflecting on a question that came up and was never fully resolved: What do we actually hire senior leaders to do, once AI can do much of what we currently describe as their job?
The honest answer, I think, is that we hire them for the things that remain irreducibly human. The judgment calls that don’t resolve cleanly into data. The representation of the organisation’s values in moments when those values are under pressure. The creation of conditions in which other people can do their best work. The willingness to be accountable — genuinely, not rhetorically — when things go wrong.
None of that is diminished by AI. If anything, it becomes more important as the technical and analytical layers of executive work get automated.
The New CXO isn’t a different species of executive. They’re a more essential version of what good executives have always been.
I co-chaired this session with my colleague Steven at EGN Singapore’s Executive Leadership peer group. The reflections here are my own synthesis; the session itself was held under Chatham House rules.