When AI Joins the Team

When we wrote The End of Leadership As We Know It, we argued that the command and control playbook inherited from the twentieth century had stopped working. Organizations had become too interconnected, too fast moving, and too unpredictable for any single leader to direct from the top. What we did not fully anticipate was how quickly the workforce itself would begin to change. Today, senior leaders are not simply managing people who use AI. Increasingly, they are leading teams in which AI agents do meaningful work alongside human colleagues.

This is no longer speculative. MIT Technology Review reports that early deployments of agentic AI in functions like customer service, HR, and sales are already producing productivity gains of 30 to 50 percent, and that the autonomy of these systems positions them more as collaborators than as tools. Researchers at PwC describe a parallel shift in role design, away from narrow specialist jobs and toward broader, outcome-focused roles as agents absorb specialized tasks. Whatever your industry, some version of the hybrid human and AI team is coming to you, probably sooner than your strategic plan assumes.

The old instincts will fail here first

Faced with something this new, most leaders reach for familiar instincts: get in control, build the detailed plan, wait for certainty before committing. We spent an entire book describing why those instincts fail in complex environments, and hybrid human and AI teams are about as complex as environments get. The technology changes monthly. The most valuable use cases emerge from the front lines, not the corner office. Nobody, including the vendors, can tell you exactly what will work in your context.

There is also a subtler trap. Agents are compliant. They do not push back, get tired, or resign. For leaders who never fully gave up the psychological need for control, a workforce of tireless digital subordinates can feel like vindication of the old model. It is not. The hard problems in this transition are human problems: trust, judgment, role clarity, and the willingness of people to redesign their own work. Those problems do not yield to tighter control. They yield to leadership.

What actually predicts success

The best evidence available points in the same direction as everything we have learned about leading through complexity. Microsoft's 2026 Work Trend Index found that employees whose managers openly use AI themselves, set clear quality standards for AI-assisted work, and create psychological safety around experimentation report substantially higher AI readiness and are far more likely to use these tools at a high level. Read that finding carefully. The differentiator is not the technology stack, the vendor, or the budget. It is leadership behavior.

Stanford's research on the future of work with AI agents reaches a complementary conclusion: as agents take on more of the information processing that once filled our days, the premium shifts to interpersonal and organizational skills. The capabilities that traditional leadership development dismissed as soft are becoming the hard currency of the AI era. This is precisely the shift we describe at EdgePoint when we talk about how exceptional leaders think. Seeing systems, making sense of ambiguity, and mobilizing others were always the real work of leadership. AI is simply stripping away everything that obscured that fact.

It helps to remember that leadership has absorbed shocks like this before. The rise of systematic management a century ago forced leaders to trade personal control for organizational control. The human relations movement forced them to take the psychology of knowledge workers seriously. Each transition felt disorienting to the generation living through it, and each produced a new set of practices suited to a new kind of organization. This transition will too. The difference is speed: leaders do not have decades to let a settled playbook emerge. They are writing it in real time, inside their own organizations, whether they intend to or not.

Three moves for leading hybrid teams

See the system, not the software. The organizations capturing real value are not bolting agents onto existing workflows. They are rethinking how work flows: which outcomes matter, which tasks agents should absorb, and where human judgment must remain non-negotiable. That is a leadership question, not an IT question, and delegating it entirely to a transformation office is a way of avoiding it. Part of seeing the system is making the division of labor explicit. For critical roles, spell out how AI should be used, what must remain human, and which skills now matter most. Ambiguity here breeds both reckless automation and quiet resistance; clarity breeds trust.

Lead the experiment visibly. You do not need to be the most technical person in the room. You do need to make learning safe. Use the tools yourself, in view of your team. Reward intelligent experiments that fail, not just initiatives that succeed. In complex environments, the organization that learns fastest wins, and people take their cues about whether learning is safe directly from you.

Reinvest the dividend in what is distinctly human. When agents reclaim hours from your calendar and your team's, decide deliberately where that capacity goes. Left unmanaged, it evaporates into more meetings and more email. Redirected with intention, it becomes time for the activities that compound: developing people, deepening relationships, and thinking clearly about what comes next. One caution belongs here as well. The tasks agents absorb first are often the same tasks on which younger professionals have always built judgment. If you automate the apprenticeship away without building something deliberate in its place, you will feel it in your leadership pipeline within a few years.

Feeling fine, again

We closed our book by borrowing a line from R.E.M. and saying that we feel fine about the end of leadership as we know it. We still do. AI agents will take over a great deal of what managers used to do. They cannot take over leading. Seeing the whole system, making sense of ambiguity, mobilizing people around what matters: that work belongs to humans, and there is more of it ahead than ever.

The leaders who thrive in this transition will be the ones who let go of control a little more, get curious a little faster, and build the human trust that determines whether any transformation actually sticks. That has been the direction of travel for twenty years. The agents are just accelerating the timetable.