Adaptive Leadership for Hybrid Human-AI Teams (Part 3 of 3): Extending the Framework for the AI Era
Part 1 of this series revisited the classic technical-versus-adaptive distinction. Part 2 looked at why hybrid human-AI teams break several of Adaptive Leadership's founding assumptions: about time, about who counts as a participant in the system, about identity, and about where accountability sits. This final piece asks what to do about it, both in how the framework itself needs to extend, and in the practical work leaders now have in front of them.
1. "Give the work back to the people," reinterpreted
This principle becomes both more subtle and more important. The risk in AI-enabled teams isn't that the work goes undone, it's that the work gets handed to the machine instead of the people, in a well-intentioned effort to reduce human effort. Adaptive Leadership's response is to design systems where humans must interpret outputs rather than just accept them, teams are expected to challenge AI conclusions, and responsibility stays traceable to a person. Used well, AI doesn't remove the adaptive work. It reallocates human effort toward it.
2. Change becomes recursive, not linear
Classic adaptive work unfolds through cycles of experimentation and learning. AI systems add a layer on top: model drift, retraining cycles, continuous updates, shifting guardrails. The system itself keeps changing while people are still changing in response to it. Adaptive work becomes co-evolution rather than a single arc from disruption to resolution.
3. Authority becomes unstable
Traditionally, authority in organisations has tracked a simple formula: expertise plus position. AI complicates that formula because it doesn't respect either variable evenly. In some narrow domains AI now matches or exceeds expert-level performance, and there's growing field evidence that AI access narrows the gap between less experienced and more experienced knowledge workers, rather than simply amplifying whoever was already strongest (Dell'Acqua et al., 2023). A junior team member with strong AI fluency can, on some tasks, produce work close to what a senior colleague would produce alone.
That doesn't mean expertise stops mattering. It means authority can no longer be assumed to track seniority automatically, and that detachment creates status threat, resistance, hidden sabotage, and decision paralysis in teams that haven't named it. The leadership task is to help the system renegotiate authority without collapsing trust in the process. That's adaptive work in its purest form, and it's mostly absent from mainstream AI advice.
4. Regulating distress in an AI-accelerated system
Heifetz's model has always centered the idea of a productive zone of disequilibrium: enough discomfort to drive learning, not so much that the system fractures (Heifetz, 1994). AI pushes systems toward both edges of that zone at once, into overload from too much change too fast, and into disengagement as people feel replaced or irrelevant.
Leaders now have to regulate cognitive overload from too many outputs and options, identity threat as people ask what their role even is anymore, and trust erosion as people question whether they can rely on the system in front of them. The counterintuitive move is to slow down interpretation even as execution speeds up, which runs directly against the grain of most AI advice.
5. Orchestrating conflict: human versus machine judgment
A new kind of conflict is showing up inside teams: human intuition against model output, ethical discomfort against statistical optimisation, experience against probabilistic reasoning. Most organisations try to suppress this friction. Adaptive Leadership does the opposite and surfaces it deliberately, for instance by requiring teams to articulate when they disagree with AI, making override decisions explicit and discussable, and treating disagreement as data rather than failure.
6. From "teams" to "adaptive systems"
This is the biggest conceptual shift. Most AI leadership thinking still assumes teams plus tools. A Heifetz-style model instead sees a living adaptive system, with humans and AI as interacting agents. Leadership shifts from managing people to designing the conditions for ongoing adaptation: feedback loops, decision rights, learning cycles, and error visibility.
Extending Adaptive Leadership for the AI era
To remain relevant, Adaptive Leadership needs to expand in four directions.
From mobilising people to mobilising hybrid systems. Leaders need to diagnose interactions between humans and AI agents, understand algorithmic behaviour as part of the system, design workflows where humans and AI co-shape outcomes, and anticipate how AI-generated signals influence human behaviour. This isn't about treating AI as a colleague. It's about recognising its systemic impact.
From holding steady to regulating continuous disequilibrium. The work is no longer about managing heat in discrete moments. It's about sustaining psychological, cognitive, and emotional capacity in an environment where change doesn't stop. That means monitoring cumulative load, building recovery cycles, designing humane pacing, and creating rituals that restore coherence. This is adaptive work at the level of the nervous system, not just the org chart.
From values and loyalties to values, loyalties, and guardrails. Adaptive Leadership now has to integrate AI governance, transparency, contestability, ethical boundaries, human override rituals, and clarity about what AI is for and not for. Leaders are navigating competing values and competing algorithmic logics at the same time.
From loss and meaning-making to identity and dignity in hybrid teams. Leaders need to help people renegotiate identity in the presence of automation, reclaim agency when systems feel more powerful than they are, articulate what remains uniquely human, and build narratives of dignity, contribution, and purpose. This is where systemic coaching, trauma-informed practice, and cultural intelligence stop being optional extras and become essential leadership skills.
Extending the lineage
If you take one thing from this series, take this: AI increases the amount of adaptive work in a system, it doesn't decrease it. Most organisations get this exactly backwards. Adaptive Leadership remains one of the most powerful frameworks we have for navigating complexity. But the world it was built for, one of human-only teams, human-paced change, and human-generated meaning, no longer fully exists. Hybrid human-AI teams require an evolution of the framework, not a rejection of it. Leaders now face adaptive challenges involving distributed agency, algorithmic influence, identity disruption, continuous change, ethical ambiguity, and systemic co-evolution, all at once. The danger was never that AI would replace human leadership. The danger is that organisations will use AI as a way to avoid leadership altogether.
A simple operating model can help you find where to start
1. Diagnose — where is AI creating adaptive pressure, not just efficiency gains?
2. Expose — where are people avoiding judgment by deferring to AI?
3. Redistribute — what work must stay irreducibly human?
4. Design tension — where should AI and humans be intentionally disagreeing?
5. Regulate pace — where do you need to slow down thinking, even with faster tools?
This is the next chapter in the Adaptive Leadership lineage. It's a chapter still waiting to be written, and it's one leaders, coaches, and scholars need to write together.
If this series has raised more questions than it's answered for your own organisation, that's rather the point. I'd welcome a conversation about where to start. You can book a 30-minute session with me directly: https://calendly.com/tara-minchin-aintreeleadership/30-minute-session
References
Dell'Acqua, F., McFowland, E., III, Mollick, E. R., Lifshitz-Assaf, H., Kellogg, K., Rajendran, S., Krayer, L., Candelon, F., & Lakhani, K. R. (2023). Navigating the jagged technological frontier: Field experimental evidence of the effects of AI on knowledge worker productivity and quality (Working Paper No. 24-013). Harvard Business School.
Heifetz, R. A. (1994). Leadership without easy answers. Harvard University Press.