Posted by 
If leaders use AI only to optimise performance, they risk building highly intelligent organisations that have forgotten how to rethink. Real innovation requires single-, double-, and triple-loop learning.
Artificial intelligence may become the most powerful management tool of this decade. It may also become the most efficient way to kill ingenuity.
That is the paradox. At the very moment leaders are being told that innovation depends on ingenuity, many organisations are using AI to tighten their grip on yesterday’s logic. They are getting better at prediction, optimisation, and control, but not necessarily better at innovation.
The problem is not the technology. The problem is the learning model around it.
Most AI is deployed inside what Chris Argyris called single-loop learning.
Single-loop learning is about correction within an existing frame. Something goes wrong, the system adjusts, and performance improves. Sales soften, the pricing engine recalibrates. Inventory drifts, the forecast model retrains.
This is where AI is genuinely powerful. It can identify patterns, process huge amounts of information, and improve decisions at a speed and scale no management team can match. If the goal is to do things right, AI is exceptional.
But ingenuity rarely begins with doing the current thing better.
The problem with single-loop learning is not that it is incomplete.
A company can become brilliant at improving performance inside an old model while missing the fact that the model itself is failing. AI can help organisations move faster down a path they should have left behind. It can make stale strategy look scientific.
This is why so many companies feel digitally advanced and strategically tired at the same time. They have more intelligence in the system, but less challenge in the room.
Real innovation begins with double-loop learning.
Double-loop learning asks a harder question. Not “How do we improve this?” but “Why are we doing this in the first place?” Not “How do we hit the target?” but “Why is this the target?” It tests the assumptions, norms, and mental models behind action.
That is where real novelty enters. New products, new strategies, and new business models rarely come from optimisation alone. They come from reframing. They come from leaders willing to ask whether the current definition of value is too narrow, whether the current customer problem is the right one, or whether the current operating logic belongs to a world that no longer exists.
AI does not naturally do this. It is trained on historical data, tuned to current KPIs, and embedded in legacy processes. It can improve the frame, but it does not naturally question the frame.
But even double-loop learning is not enough.
Triple-loop learning asks the deepest question of all: how do we decide what matters here, and what kind of organisation are we becoming through those decisions?
Single-loop learning improves action. Double-loop learning challenges assumptions. Triple-loop learning redesigns the system that produces both. It looks at culture, incentives, leadership habits, meeting structures, performance measures, and power dynamics. It asks whether the organisation knows how to learn, or whether it only knows how to comply.
This matters because ingenuity does not appear by accident. It is shaped by environment. If meetings reward certainty over curiosity, ingenuity shrinks. If incentives reward short-term efficiency over experimentation, ingenuity shrinks. If leaders use AI to close debate rather than open it, ingenuity shrinks.
Triple-loop learning asks whether the organisation knows how to learn, or whether it only knows how to comply. As Drucker observed, “So much of what we call management consists in making it difficult for people to work.” Triple-loop learning is the discipline that names that problem and starts dismantling it.
That is why the Drucker Forum’s 2026 theme matters so much. If innovation is to become an enduring capability, then leaders have to build systems that keep ingenuity alive, not just systems that keep performance stable.
The biggest risk of the AI era is not simply automation. It is the combination of intelligent tools with shallow learning.
An organisation can have excellent AI and still become less innovative. It happens when teams become better at hitting targets but less willing to question them. It happens when dashboards become more sophisticated but conversations become less honest. It happens when speed replaces reflection.
In those conditions, AI does not create innovation. It merely accelerates conformity.
If leaders want AI to strengthen ingenuity rather than suppress it, they need to work in all three loops.
Let AI do what it does best: pattern detection, forecasting, operational optimisation, scenario analysis, and decision support. Use it to improve execution. That is real value.
Force second-loop conversations
Create time to question assumptions. Ask which metrics no longer serve the business. Ask what beliefs are baked into the current strategy. Ask which customer needs are emerging outside the current frame. If these questions are not being asked, innovation is already weakening.
Look beyond the tool and into the management system. Audit incentives, reporting lines, meeting formats, approval processes, and leadership behaviour. Ask whether the organisation genuinely rewards curiosity, challenge, and experimentation, or merely says it does.
In major decisions, let people articulate their reasoning before they see the model’s answer. This reduces the risk that the machine silently narrows the conversation before human judgement has even begun.
Most organisations reward polished answers. Innovative organisations also reward reframing. If nobody gets recognised for asking the question that changes the conversation, the culture will drift back to safe optimisation.
The most important leadership choice in the AI era is not whether machines will replace people. It is whether organisations will use AI to deepen learning or to avoid it.
AI is already very good at first-loop learning. Human beings remain essential to second-loop learning. Leadership’s hardest task is third-loop learning: building the organisational conditions in which ingenuity can survive pressure, scale across teams, and endure beyond individual talent.
That is the real management challenge now.
When everything depends on ingenuity, the winners will not be those who automate the fastest. They will be those who can improve performance, challenge assumptions, and redesign the system in which learning happens.
The decisive question for every executive team is simple: Are we using AI to think for us, or to force us to think better?
About the author:
Nick Hixson is a business advisor and writer on strategy and leadership. He explores how complexity and human behaviour shape organisations. He is a Peter Drucker Associate and chairs the Advisory Board of the World Institute for Action learning.