It’s Not a Tech Problem: What Adaptive Leadership Knows About Making AI Work

By Dana Houston Jackson, Founder and Principal Change Architect, Change – With a Coach Approach, LLC
The Inherited Operating System for Organizations – Expired Assumptions | Part 6
Most AI rollouts are run like software installs. The hard part was never the software (at least it wasn’t until agentic AI went wild at Hugging Face).
In a nutshell – what we learned (right or wrong)
The big idea: Some problems have known fixes. Others ask people to learn, let go of something and change how they work. Leadership means telling them apart.
What’s expired: Treating every change, AI included, as a technical problem an expert can solve and hand down.
What replaces it: Adaptive leadership: get on the balcony, give the work back to the people who have to change, and hold steady while they learn.
Here’s a scene you may recognize. AI tokens gone by the 10th. Everyone’s using AI solo. The licenses are bought, the training is done, and the work itself looks exactly like it did last year. The real work keeps going home, to personal accounts where nobody’s counting tokens, and leadership is asking why the investment isn’t paying off.
If that sounds familiar, this article is for you. The first five parts of this series looked at the thinkers who wrote the inherited operating system. This one looks at a group of thinkers who explain why it struggles with change, and AI is the biggest change most of us will lead. It covers what adaptive leadership is, why it got watered down, where AI clashes with the way we usually lead change and what to do instead. And as usual in all my articles, it ends with five questions and one small step. Because what’s the use of writing nowadays if it isn’t read, and if agreed with, then use attempted…
The world they were solving for
By the early 1990s, many leaders were finding that the old playbook didn’t work on the problems in front of them. Ronald Heifetz, a psychiatrist by training who went on to found the Center for Public Leadership at Harvard Kennedy School, kept seeing the same pattern in government, business and community life. People in charge were expected to have the answer. On the hardest problems, nobody did.
His answer, in Leadership Without Easy Answers (1994), was a simple distinction that still does a lot of work. Technical problems can be solved with know-how we already have. An expert can fix them. Adaptive challenges can’t be solved that way, because the people with the problem have to learn something new, give something up or change what they value. Nobody can do that for them.
What they built
Heifetz and his colleagues turned that distinction into a practice. Leadership on the Line (2002), with Marty Linsky, and The Practice of Adaptive Leadership (2009), with Linsky and Alexander Grashow, laid out the moves:
- Get on the balcony. Step out of the action long enough to see the pattern, not just the next fire.
- Diagnose before you act. Ask whether this is technical, adaptive or both. Most real change is both.
- Give the work back. Don’t take the problem off people’s shoulders. Help them solve it, because they’re the ones who have to change.
- Regulate the heat. Keep the pressure high enough that people engage and low enough that they don’t shut down.
- Hold steady. People push back on whoever asks them to change. Expect it, and don’t take it personally.
The warning at the heart of their work: one of the most common ways leaders fail is treating adaptive challenges as if they were technical problems.
Others built on the same ground. Robert Kegan and Lisa Lahey, first in Harvard Business Review (2001) and then in Immunity to Change (2009), show why smart people don’t change even when they sincerely want to: a hidden competing commitment protects something they care about. Mary Uhl-Bien and her colleagues’ complexity leadership theory (2007) argued that leadership models of the last century were products of top-down, bureaucratic thinking, built for physical production, not knowledge work. Dave Snowden and Mary Boone’s Cynefin framework (2007) gave leaders a way to tell a complicated problem (an expert can analyze it) from a complex one (you have to probe, learn and adapt). Jennifer Garvey Berger and Keith Johnston’s Simple Habits for Complex Times (2015) turned all of this into everyday habits: ask different questions, take multiple perspectives and see the system.
Who's carrying it forward now
Adaptive leadership isn’t a 1990s museum piece. The most interesting work on it is happening right now.
- Sharon Ravitch and Liza Herzog. Leadership Mindsets for Adaptive Change: The Flux 5 (Routledge, 2024) names five mindsets leaders need when change won’t sit still: inquiry, humanizing, systems, entrepreneurial and equity. It’s built for practice, with exercises for leaders and teams.
- Amy Edmondson. Right Kind of Wrong (2023) separates intelligent failure, the kind you learn from in new territory, from the preventable kind. AI adoption is new territory for everyone.
- Margaret Heffernan. Uncharted (2020) argues that we have to stop expecting the future to be predictable and build the capacity to adapt instead, through experiments and scenarios rather than one fixed plan.
- Jennifer Garvey Berger. Her work on leading in complexity keeps turning adaptive ideas into daily habits: ask different questions, take multiple perspectives, see the system.
- Dave Snowden. He has kept developing Cynefin, which today calls the obvious domain “clear,” into tools for sense-making in messy situations.
- Ronald Heifetz and Marty Linsky. Cambridge Leadership Associates, the firm they co-founded in 2002, still teaches adaptive leadership to organizations around the world.
Practitioners keep it honest. Dr. Ross Wirth, retired Dean of the College of Business at Franklin University, brings decades of experience across the energy industry, academia and consulting. With Dr. Reg Butterfield, he describes a purpose-driven organization built for routine, adaptive change that everyone leads, freed from bureaucratic controls. In a 2025 episode of the Value Creators podcast, “From Structure to Flow: How Organizations Evolve Beyond Industrial-Era Mindsets,” he makes the case, as best I can tell from the episode and its notes, that change run as a one-off project is doomed, that older structures suffocate adaptability and that organizations need to look hard at radical decentralization. His LinkedIn headline says it plainly: “The organizational change profession is being disrupted. Learn how to evolve before you become irrelevant.” That’s adaptive leadership turned on its own field.
How it got handed down
Adaptive leadership reached most of us the way Fayol’s ideas did: as a summary of a summary.
- 1994: Heifetz publishes Leadership Without Easy Answers.
- 2002: Leadership on the Line, with Marty Linsky, makes the ideas practical for managers.
- 2007: Complexity leadership theory and Cynefin give the same insight a research base and a decision tool.
- 2009: The Practice of Adaptive Leadership adds tools and worksheets. Immunity to Change explains why change stalls inside people, not just systems.
- 2010s onward: “Adaptive” becomes a competency on a leadership model, often next to “agile” and “resilient.” Nothing says “learning organization” like a word on a poster.
- 2020–2024: A pandemic, then generative AI. Heffernan, Edmondson, and Ravitch and Herzog bring the ideas back to practice for a world nobody can predict.
- Today: Many leaders can say “adaptive challenge” and still run AI adoption as a project plan with a go-live date.
The full title of the upcoming book Leading a Five-Generation Workforce: What Each Generation Brings and What Adaptive Leaders Must Understand names the gap. In it, we write: “The central discipline is learning not to confuse what was once adaptive back then with what remains adaptive today.” What got handed down was the vocabulary. What got lost was the hard part: the leader doesn’t get to solve it for people.
Where it still shows up
- AI rollouts owned by IT and measured by licenses and logins.
- Change plans with a communications calendar and no time for people to work out what changes for them.
- Leaders who announce the answer, then call the pushback “resistance.”
- “Lessons learned” sessions held after the project is over, when nobody can use them.
- Pilots that succeed in one team and never spread, because the next team was told, not involved.
What it costs us now
AI is close to a perfect adaptive challenge disguised as a technical one. The technical part is real: tools, data, security, integration. But the part that decides whether it works is adaptive. People have to rethink what their expertise is worth, which tasks they hand over, what “good work” means now and who they are at work when the routine part of their job goes away. No vendor can install that.
When leaders treat it as technical, they get compliance on paper and workarounds in practice. They get people who hide their AI use, or hide from it. And they spend a lot of money proving that a tool works in a pilot and doesn’t work in the organization. That dog won’t hunt.
The bigger picture: Taylor, Weber and Fayol still shape the default operating system most organizations fall into: specialization, hierarchy, rules, measurement and management as control. That system is built for technical problems. It assumes the answer lives at the top and travels down. As AI agents take over specialized tasks, the questions left for people are mostly adaptive ones, starting with “what do I do now that my specialty is changing?” An operating system built to hand down answers can’t help people learn their way into a new role. Adaptive leadership can.
What to do instead
Start every AI decision by asking which part is technical and which part is adaptive. Then lead each one differently.
- Fix the technical part with experts. Data, security, access and integration need people who know how. Don’t make the front line figure out single sign-on.
- Give the adaptive part back to the people doing the work. Let teams decide which tasks AI takes, which stay human and what “good” looks like now. This is Follett’s power with, in practice.
- Make it safe to learn out loud. Amy Edmondson’s research on psychological safety is the holding environment Heifetz describes: people can say “I don’t know how to use this yet” or “this didn’t work” without paying for it.
- Surface the competing commitments. Use Kegan and Lahey’s question: what are we committed to that keeps us from changing? Often it’s protecting expertise, status or job security. Name it and talk about it.
- Probe, then scale. In complex situations, Cynefin says run small, safe-to-fail experiments before you write the policy.
This is the heart of a coach approach. Coaches don’t hand people answers to adaptive challenges. They ask the questions that help people find their own, and they stay with them while it’s uncomfortable.
Unlearn to make room
Most of us were taught to lead change by having the plan. Here’s what’s worth unlearning.
| What many of us were taught | Why it's expired | Make room for |
|---|---|---|
| The leader's job is to have the answer. | On adaptive challenges, nobody has it yet, including the leader. | Leaders who frame the question and give the work back. |
| Pushback means resistance. | Pushback is often information about what people stand to lose. | Asking what people are protecting and what would help. |
| A change is done at go-live. | Adaptive change happens after go-live, as people learn. | Measuring what people can do differently, not what was launched. |
Paul Cheek’s No One Works Here describes decision cycles shrinking from quarters to hours as AI spreads. An organization can’t learn at that speed if every adaptive question has to wait for an answer from the top. It means redesigning who gets to work things out, not just training people on the tool.
How adaptive leadership fits with the others: Taylor, Weber and Fayol built for technical problems: find the one best way, write the rule, check the work. Follett and McGregor saw the human side. Adaptive leadership explains why that human side decides whether change sticks. Heifetz is Follett’s law of the situation with a toolkit.
Where AI meets adaptive leadership
Where it clashes: The inherited operating system treats AI as a technical rollout: pick the tool at the top, train everyone, track usage. That ignores the adaptive part entirely, so people protect what they’re afraid of losing. Snowden would say it’s managing a complex problem as if it were merely complicated. Ross Wirth’s warning applies to change teams too: a change function that only knows how to run a project plan will be disrupted along with everyone else.
Where it helps: AI can be a great balcony. It can show patterns across teams, summarize what people are struggling with and free up time for the conversations adaptive change needs. Used well, it gives leaders more room to think, not less.
What has to be redesigned: Split every AI initiative into technical work (owned by experts) and adaptive work (owned by the teams who have to change). Give teams real decision rights over how AI fits their work. Build learning time into the job. Train leaders and change practitioners in adaptive skills, not just project plans, before the rollout, not after.
Check your own organization
Picture a community hospital that rolls out an AI “ambient scribe” in 2025. It listens to the visit and drafts the clinical note, so doctors stop charting at the kitchen table at 10 p.m. The technical work goes well: the notes are accurate, the system is secure and the vendor’s dashboard shows strong adoption. Six months in, the picture is mixed. Some physicians love it. Others sign the notes without reading them closely. Nurses say the drafts miss what they flag at the bedside, and nobody is sure who owns a mistake in a note the AI wrote and a doctor signed. Leadership’s first instinct is a refresher training and an adoption target. The better move is the balcony. This isn’t a software problem. It’s a question about professional judgment, accountability and trust, and only the people doing the work can answer it. So the hospital gives the work back: a physician, a charge nurse, a coder and a patient-safety lead get one unit and six weeks to decide what “a good note” means now, who reviews what, and when the AI draft must not be trusted. Their answers become the policy, and the policy finally sticks.
- Which of your AI efforts are being treated as purely technical?
- What would people have to give up or learn for AI to really work on your team?
- Where is “resistance” actually information you haven’t asked about?
- Who owns the adaptive work in your AI rollout, if anyone?
- How safe is it on your team to say “I don’t know how to use this yet”?
One small first step: Take one AI initiative and draw a line down a page. On the left, list the technical work. On the right, list the adaptive work. Then ask: who owns the right-hand column? If the answer is “nobody” or “IT,” give it back to the team that has to change, and give them time to work on it.
Next in the series: today's big thinkers on what AI transformation actually requires, and the newer management ideas now in the way.
Quick Answers
What is adaptive leadership?
Developed by Ronald Heifetz and colleagues, adaptive leadership separates technical problems, which experts can fix, from adaptive challenges, which require people to change their habits, beliefs and ways of working.
Why is AI adoption an adaptive challenge?
The technical part is real: tools, data, security and integration. But what decides whether AI works is adaptive: people rethinking what their expertise is worth, which tasks to hand over and what good work means now.
How do leaders lead AI as an adaptive challenge?
Ask which part of each AI decision is technical and which is adaptive. Fix the technical part with experts, and give the adaptive part back to the people doing the work.
The Inherited Operating System series
- Kickoff: Your Organization Is Running on 100-Year-Old Code
- Part 1: Who Gets to Think? Frederick Taylor
- Part 2: The Rule Is Safer Than Your Judgment. Max Weber
- Part 3: Management Means Control? Henri Fayol
- Part 4: Power With, Not Power Over. Mary Parker Follett
- Part 5: Theory X, Theory Y. Douglas McGregor
- Part 6: It’s Not a Tech Problem. Adaptive Leadership (you are here)
- Part 7: The New Code Is Already Being Written
- Part 8: The Way Forward
Sources
Berger, Jennifer Garvey, and Keith Johnston. Simple Habits for Complex Times: Powerful Practices for Leaders. Stanford, CA: Stanford Business Books, 2015.
Cheek, Paul. No One Works Here: How AI-Driven Enterprises Are Raising the Bar for Business, Leadership, and Competition. Hoboken, NJ: Wiley, 2026.
Edmondson, Amy C. “Psychological Safety and Learning Behavior in Work Teams.” Administrative Science Quarterly 44, no. 2 (1999): 350–383.
Edmondson, Amy C. Right Kind of Wrong: The Science of Failing Well. New York: Atria Books, 2023.
Heffernan, Margaret. Uncharted: How to Navigate the Future. New York: Avid Reader Press, 2020.
Heifetz, Ronald A. Leadership Without Easy Answers. Cambridge, MA: Belknap Press of Harvard University Press, 1994.
Heifetz, Ronald A., and Marty Linsky. Leadership on the Line: Staying Alive Through the Dangers of Leading. Boston: Harvard Business School Press, 2002.
Heifetz, Ronald A., Alexander Grashow, and Marty Linsky. The Practice of Adaptive Leadership: Tools and Tactics for Changing Your Organization and the World. Boston: Harvard Business Press, 2009.
Houston Jackson, Dana, et al. Leading a Five-Generation Workforce: What Each Generation Brings and What Adaptive Leaders Must Understand. Forthcoming, April 2027.
Kegan, Robert, and Lisa Laskow Lahey. “The Real Reason People Won’t Change.” Harvard Business Review, November 2001, 85–92.
Kegan, Robert, and Lisa Laskow Lahey. Immunity to Change: How to Overcome It and Unlock the Potential in Yourself and Your Organization. Boston: Harvard Business Press, 2009.
Netflix. Instadocs: AI Gone Wild (documentary special, Words + Pictures). Released October 12, 2026. netflix.com
Snowden, David J., and Mary E. Boone. “A Leader’s Framework for Decision Making.” Harvard Business Review, November 2007.
Uhl-Bien, Mary, Russ Marion, and Bill McKelvey. “Complexity Leadership Theory: Shifting Leadership from the Industrial Age to the Knowledge Era.” The Leadership Quarterly 18, no. 4 (2007): 298–318.
Ravitch, Sharon, and Liza Herzog. Leadership Mindsets for Adaptive Change: The Flux 5. New York: Routledge, 2024.
Wirth, Ross. Academia.edu profile. franklin.academia.edu
Wirth, Ross. “From Structure to Flow: How Organizations Evolve Beyond Industrial-Era Mindsets.” Value Creators podcast, episode 75, hosted by Hunter Hastings, October 9, 2025. buzzsprout.com
Wirth, Ross. LinkedIn profile. linkedin.com