No matter how sophisticated artificial intelligence becomes, the decisions that matter most in business still require something the machine does not possess: lived experience, moral judgment, and a feel for context. The practical question for leaders is not whether intuition beats AI, but when. The answer, drawn from experts in data science, leadership, and cognitive science, is that AI is reliable where data is complete and problems are structured, while human judgment becomes indispensable as ambiguity, novelty, and ethical stakes rise.
The Indispensable Role of Human Intuition in the Age of AI
Intuition has a reputation problem. In a data-driven society, relying on a hunch can seem outdated, yet IMD's leadership analysis notes that it "remains essential for effective decision-making." The label conceals something rigorous: intuition is learned pattern recognition, the compressed product of experience. Citing research by the Economist Intelligence Unit from 2014, IMD reports that C-suite executives split roughly three ways on important decisions—30% lean on intuition and experience, 29% on data and analytics, and 28% on advice from others. Even at the top of the data era, judgment never surrendered.
What has changed is that AI can now externalize reasoning at scale. Maria F. Martinez Vazquez, writing in an essay for the Harvard T.H. Chan School of Public Health, observes that what once stayed private and tacit in human thought is now "visible, editable, and scalable." That shift makes intuition newly valuable rather than obsolete. She quotes cognitive scientist Murray Shanahan of DeepMind, who describes large language models as "compressed mirrors of human discourse"—they reflect patterns of communication without the embodied grounding that gives human thought its texture. A mirror can show you the pattern; it cannot tell you whether the pattern means anything.
Josh Bersin makes a complementary point about interpretation. Even when a business decision seems purely logical and data-driven, he argues, "we each interpret data differently and our experience and human nature define our reaction." He cites the familiar executive meeting where the same revenue figures lead one leader to celebrate success and another to demand faster growth. The data did not change; the humans did. That interpretive layer is precisely what AI does not supply.
A Decision Framework: When to Trust AI, When to Trust Your Gut
The strongest argument for keeping humans in the loop comes from the people who build the models. An analysis from Two Sigma, the quantitative investment firm, states plainly that human intuition is vital for "assessing whether the available data can support the creation of a reliable and robust machine learning model"—and, crucially, for deciding against using machine learning at all. The firm identifies ten common scenarios where the intuitive judgment of skilled practitioners is crucial, concluding that despite technological advances, expert human input "remains indispensable."
Data quality is the first fault line. Bersin notes that the entire enterprise assumes the training data "actually has enough insights or perspectives to make this decision," and that major AI labs now admit they have "run out of data" to index, turning to synthetic data—data created by AI from data—to enrich models. When the input is thin, recycled, or biased, a confident output is not evidence of a sound recommendation.
Problem structure is the second. IMD observes that algorithms have matched or outperformed expert humans for decades on pattern recognition and quick analytical decisions over vast data. But it is "in these uncertain situations where informed intuition, based on learning and experience, comes to the fore." Harvard's essay adds a telling operational signal: engineers describe sensing when a model's output is off before the metrics say so. That early-warning capacity—detecting misalignment before it is measurable—is a practical red flag leaders can institutionalize rather than a mystical gift.
The AI-Human Decision Matrix
Use this matrix to evaluate an AI-generated recommendation before acting on it. Read each decision factor, ask the accompanying question, and let the pattern of answers determine whether to trust the machine, supplement it, or override it.
| Decision factor | Where AI is strong | Where human intuition is critical | Leader's question |
|---|---|---|---|
| Data quality & completeness | Excels when data is complete, well-structured, and representative | Judging whether available data can support a reliable model at all; spotting gaps, bias, and synthetic-data thinness | Could flawed or missing data produce this answer? |
| Problem ambiguity & novelty | Recognizes known patterns and delivers quick analytical decisions from vast data | Navigating novel, ill-defined, or uncertain situations; sensing when a recommendation is misaligned with context | Has anyone seen this situation before? |
| Ethical & strategic stakes | Processes vast information to identify potential outcomes | Moral judgment, ethics, and understanding the "why" behind a decision | Who is harmed if this answer is wrong? |
| Contextual alignment | Reflects patterns of human discourse without embodied grounding | Empathy, social relationships, foresight, and creative visioning | Does this fit what people in the room actually know? |
The matrix yields a simple decision rule. Strong data plus a structured, low-stakes problem: trust the AI's recommendation. Thin or synthetic data, a novel situation, or high moral stakes: seek more human input before acting. A recommendation that feels misaligned with context, or that experts sense is off before the metrics confirm it: override or delay. None of these triggers requires technical expertise—they require the discipline to ask.
Beyond Algorithms: Where Human Intelligence Shines
IMD identifies the specific capabilities AI currently struggles with but that come naturally to people with well-developed intuitive intelligence: "empathy and social relationships, moral judgment and ethics, foresight and visioning, creativity, invention and innovation, and taking decisions in data-poor and time-pressured situations." These are not soft extras; they are the substance of leadership.
Three stand out. First, ethical reasoning: AI's capability is limited to computational analysis, while human emotions and lived experience are integral to how people weigh harm and fairness—Bersin traces this to how feelings about a situation arise from lived experience, the stimulus itself, and even genetic makeup shaped by millions of years of evolution. Second, true creativity and strategic visioning: generating what does not yet exist, not recombining what already does. Third, deep contextual understanding: reading a room, sensing a customer's unspoken concern, or knowing which "logical" recommendation will fail for reasons no dataset captures.
IMD's conclusion is that the future of intelligence "will likely be hybrid," with analytical and intuitive human intelligence working in partnership with AI. Harvard's essay sharpens the competitive stakes: intuitive discernment of "what feels right, what fits, what aligns may confer competitive advantage when raw computational power no longer does."
Embracing the Hybrid Future: A Call to Action
Leaders should actively cultivate and integrate human intuition with AI systems, recognizing that the most robust decisions arise from a synergistic approach. This is not an argument against AI, and it is not a technical guide to building models—it is a governance discipline for using them. The measurable indicator of success is increased instances of human override or contextual adjustment of AI recommendations leading to demonstrably better outcomes, alongside a documented framework for evaluating AI outputs at critical decision points. Start by adopting the matrix above at your next consequential decision: record the recommendation, the questions asked, and whether judgment altered the outcome. The organizations that win with AI will not be the ones that defer to it fastest, but the ones that know exactly when not to.
Sources
- Intuition and AI - Vital Role - I by IMD — I by IMD - First-person Business Intelligence from the Brightest Thinkers
- Essay 2 | Intuition in the Age of AI — Harvard T.H. Chan School of Public Health
- Can AI Beat Human Intuition For Complex Decision-Making? I Think Not. JOSH BERSIN
- Why Human Intuition Is Essential in Machine Learning — Two Sigma




