Adaptive Leadership for Hybrid Human-AI Teams (Part 2 of 3): Why the Framework Must Evolve
In Part 1 of this series, I revisited Ronald Heifetz's foundational distinction between technical and adaptive challenges and argued that most AI rollouts are being managed as if they were entirely technical when what's actually being asked of people is adaptive. This piece goes a step further: hybrid human-AI teams introduce conditions the original Adaptive Leadership framework was never built to handle, and the framework itself now needs to evolve.
For more than three decades, Adaptive Leadership has offered one of the most influential frameworks for navigating complexity and change. Ronald Heifetz introduced the distinction between technical and adaptive work in Leadership Without Easy Answers (Heifetz, 1994), and later extended it into a practical toolkit with Alexander Grashow and Marty Linsky in The Practice of Adaptive Leadership (Heifetz et al., 2009). Heifetz has argued that a leader can operate with little more than a well-placed question (Heifetz, 1994). The work was never about having the answer. It was about helping a system find its own.
Most AI leadership advice asks, "How do we use AI effectively?" That's a technical question. The more useful one, and the one this series is built around, is: what is the adaptive challenge that AI is exposing in this system?
The emergence of hybrid human-AI teams introduces conditions the original framework never had to consider. AI is no longer simply a tool that supports human work. It is becoming a system actor, generating tasks, shaping workflows, influencing decisions, and altering the emotional and political dynamics of organisations. The adaptive challenges of the AI era aren't just larger or faster versions of what came before. They are categorically different, and they require an evolution of Adaptive Leadership itself.
So what are those challenges?
1. AI is not just a tool, it's a pressure test
Introducing AI into a team, surfaces exactly the fault lines Adaptive Leadership was designed to diagnose: skill obsolescence, authority confusion, hidden value conflicts between efficiency and judgement, and an overreliance on expertise at precisely the moment learning is most needed.
2. AI compresses time beyond human adaptive capacity
Adaptive Leadership assumes people need time to metabolise loss, renegotiate identity, and experiment with new behaviours. AI collapses these timeframes. Workflows shift overnight. Roles dissolve in weeks. Teams experience stacked micro-changes with little recovery between them.
The disequilibrium is no longer episodic. It is continuous. Leaders must now regulate not just the heat of a single adaptive challenge, but the cumulative load of ongoing, AI-driven transformation.
3. AI becomes a participant in the system, not a neutral tool
Adaptive Leadership is built on the premise that humans generate meaning, hold values, resist change, and interpret signals. Hybrid teams complicate that premise. The classic distinction between technical and adaptive work still holds, but it sharpens, because AI agents now generate work, prioritise tasks, shape information flows, and create new forms of ambiguity and dependency.
Technical work is becoming increasingly AI-dominant: data processing, pattern recognition, drafting, coding, optimisation, prediction. Adaptive work becomes more human-critical as a result: framing the problem, interpreting ambiguous outputs, exercising ethical judgment, and deciding what should be done rather than just what can be done.
Leaders must now diagnose systems in which non-human actors have agency-like effects, something the original framework never had to account for. The key shift is that AI expands the technical domain, which raises the stakes of the adaptive work that's left.
4. Identity, dignity, and loss take on new forms
Adaptive Leadership has always centered the emotional cost of change. AI introduces new identity ruptures on top of that:
What is my value if the system does this better than I do?
What remains distinctly human in my role?
How do I lead when the algorithm is more trusted than I am?
These aren't technical anxieties. They're existential, and they strike at the heart of dignity, belonging, and professional identity.
5. Accountability becomes distributed and ambiguous
Adaptive Leadership assumes authority and responsibility sit with people, and Heifetz's model has always recognised that people avoid adaptive work when they can find a way to (Heifetz et al., 2009). AI supercharges the avoidance options available. Hybrid teams introduce new escape routes:
algorithmic recommendations that outsource judgement ("the model said so"),
opaque decision pathways that let people hide behind complexity ("it's too technical to question"),
shared accountability between humans and systems that tips into over-trusting outputs to avoid conflict, and
the use of AI speed to bypass deliberation altogether.
This isn't a hypothetical risk. Research on automation bias has long shown that people tend to defer to automated recommendations even when they have reason to doubt them, particularly once a system has proven reliable in the past (Skitka et al., 1999; Alon-Barkat & Busuioc, 2023). Leaders now have to navigate distributed agency and algorithmic influence deliberately, and actively expose where AI is being used to avoid human responsibility, so that accountability stays traceable to a person even when a decision has been shaped by a machine.
Next week, in Part 3, I'll look at what this means for authority, conflict, and the practical shift from managing teams to designing adaptive systems, along with a simple operating model for diagnosing where the real adaptive work is hiding in your organisation.
If you're seeing these dynamics play out in your own team right now, I'd welcome a conversation. You can book a 30-minute session with me directly: https://calendly.com/tara-minchin-aintreeleadership/30-minute-session
References
Alon-Barkat, S., & Busuioc, M. (2023). Human–AI interactions in public sector decision making: "Automation bias" and "selective adherence" to algorithmic advice. Journal of Public Administration Research and Theory, 33(1), 153–169. https://doi.org/10.1093/jopart/muac007
Heifetz, R. A. (1994). Leadership without easy answers. Harvard University Press.
Heifetz, R. A., Grashow, A., & Linsky, M. (2009). The practice of adaptive leadership: Tools and tactics for changing your organization and the world. Harvard Business Press.
Skitka, L. J., Mosier, K. L., & Burdick, M. (1999). Does automation bias decision-making? International Journal of Human-Computer Studies, 51(5), 991–1006. https://doi.org/10.1006/ijhc.1999.0252