How AI challenges CEOs: automate the past or rethink the enterprise?
This article is drawn from remarks by Alfredo Altavilla, Special Advisor BYD Europe, at the Leadership Series dedicated to students of the Executive MBAs and the Global Executive MBA.
CEOs who are using artificial intelligence to do better and faster what they were already doing have not grasped the scope of this technological revolution. They are simply automating the past, whereas the future requires rethinking the enterprise, the role of leadership, and the very way decisions are made.
Their temptation is understandable. In every phase of technological discontinuity, the first reflex of organizations is to adapt the new to existing frameworks. This has happened in the past and is happening today with artificial intelligence. But, as Peter Drucker warned, “the greatest danger in times of turbulence is not the turbulence, it’s acting with yesterday’s logic.” And that is exactly what is happening: companies adopting new tools to replicate old mental models.
The issue, then, is organizational and cultural. AI is not, and must not be, an IT project to be delegated to a specialized function: it is a transformation of leadership. It shifts the source of competitive advantage from scale to intelligence, and the lever through which value is created from assets to data. It also changes the pace of organizations: no longer long and predictable cycles, but continuous adjustments, often driven by weak signals. As Eisenhower put it, “plans are worthless, but planning is everything.”
Above all, it changes the function of the leader: from control to orchestration. For decades, leadership has been associated with the ability to make the best decisions and, based on those, tell others what to do. Today, leaders must build systems that enable better decisions at scale. This is not a loss of power, but a redefinition of it.
So, are we using artificial intelligence to automate what we already did, or to redesign the company? In the first case, AI becomes a “simple” efficiency accelerator; in the second, a driver of transformation.
Organizations that grasp this transformation are already evolving toward flatter, faster, and more adaptive models. It is no coincidence that some industrial experiences have shown how complex operations can be managed with surprisingly lean structures. Fewer hierarchical levels do not mean less control, but a different kind of control, based on richer and more timely information.
The shift is from hierarchical structures to systems driven by distributed intelligence; from functional silos to cross-functional logics, which require CEOs to make a greater effort to listen; from periodic planning to continuous optimization, enabled by constant monitoring and increasingly granular KPIs. Judgment does not disappear, but is transformed: from individual experience to data-driven decision-making.
This also entails a redefinition of skills. It is not necessary to understand every technical detail of artificial intelligence or its granularity, but it is essential to build teams capable of doing so: selecting, developing, and retaining talent able to operate in uncharted territory. Because, ultimately, the leader’s true distinctive competence remains the ability to choose the right people.
In this context, leadership ceases to be directive and becomes architectural. The leader does not decide everything, but must establish what truly matters: what to automate and what to keep human; where to place responsibility; which values should guide and constrain algorithms. In other words, leaders must define the boundaries within which machines operate and people act.
Boards are also called to raise their game. They must ask where and why they are delegating judgment to machines, define clear boundaries between human and automated decisions, and assume ethical, regulatory, and reputational responsibility for those choices. In a world where scenarios are multiplying and becoming more uncertain, strategy can no longer be a linear path, but a continuous exercise in exploration and adaptation.
This change is even deeper in industrial sectors, where AI meets capital-intensive contexts, safety-critical environments, very long asset life cycles, and complex supply chains. Here there is a risk of losing efficiency and dispersing tacit knowledge accumulated over years of experience. And it is precisely here that artificial intelligence shows one of its most significant promises: embedding experience into systems, making organizations faster without making them more fragile.
The human dimension, however, should not be underestimated. In a context where everyone is, in many ways, starting over, as many managers observe, emotional intelligence, empathy, and the ability to guide people through a potentially destabilizing change become central. CEOs must create the conditions for others to contribute to redefining the rules.
What emerges is a new leadership profile. Not technocrats obsessed with tools, nor abstract futurists. What is needed are figures who combine industrial credibility, AI literacy (including its limits), systemic thinking, and the ability to provide human and cultural guidance. Leaders capable of continuous learning, because the speed of organizations will surprise even those who lead them. Artificial intelligence will not replace industrial leaders, but it will certainly replace those who do not know how to lead with AI.
Ultimately, the challenge remains the same as ever: to take people from where they are to where they have never been. It is a definition of leadership that spans eras, but today it takes on new urgency. It is no longer enough to improve what exists; it takes the courage to question it. And perhaps it is precisely here that the ultimate quality of a leader is measured: in the impact they manage to generate over time, in their ability to challenge the status quo, knowing that leadership is never permanent, but always transitory.
Ron Bloom, a former successful banker called by the Obama administration to help restructure the American automotive industry, agreed to move from multimillion-dollar compensation to a symbolic salary. When I asked him why, he replied with disarming simplicity: one day, he would like his family to be able to write a single sentence: “Ron made a difference.” That is what every leader should aspire to, even in the age of artificial intelligence.
