Introduction

Artificial intelligence (AI) is a technology with profound transformative power. Arguably one of the most transformative technologies in human history. For the first time, we are confronted with machines that possess the technical ability (whether real or apparent) to “reason”, demonstrate creativity, and make judgments in complex situations. They can therefore make decisions autonomously.

Consequently, it is essential not only to develop a deep understanding of the nature of this technology from the perspective of its practical application, but also to engage in a moral and philosophical debate that defines its boundaries. Pope Leo XIV’s encyclical Magnifica Humanitas offers an excellent reflection and a forward-looking step on one of the defining issues of our generation: how to integrate these new technological capabilities within a genuinely human vision of progress. The encyclical warns that technology must not become an end in itself, nor should it replace moral judgment or responsibility. On the contrary, the most powerful technologies should be directed toward human flourishing, shared prosperity, and the integral development of individuals.

How does this debate translate into the heart of the corporate world, which is inherently competitive and driven by a logic of cost and benefit? In universities and business schools, this reflection has resulted in an intense discussion about how organizations will evolve as a consequence of the large-scale adoption of AI, and what corporate leadership itself will mean in a context where machines may eventually be capable of making decisions on their own. I aim to examine first the nature of AI in order to understand its implications for the reconfiguration of corporations, and to outline the new leadership skills required to achieve the goals of productivity and economic success, without sacrificing the pursuit of the common good.

The article makes three contributions, drawn on my previous research. First, it presents a conceptual model of the “AI-organization” built around a digital core, an AI- powered “reasoning” engine and a final human decision layer. Second, it introduces the notion of “cognitive advantage” as a new source of competitive advantage and productivity. Third, it develops a humanistic interpretation of AI management through a dialogue between contemporary management research and the Social Doctrine of the Church, not treating it as an external ethical constraint on technological innovation, but rather as a conceptual framework for understanding how AI can become a source of human dignity and economic development.

From computation to intuition. The early evolution of AI

Although the emergence of ChatGPT and Large Language Models (LLMs) has brought AI into the mainstream in recent years, its foundations date back more than eight decades. The pioneering work of McCulloch and Pitts on artificial neurons, together with Alan Turing’s groundbreaking ideas, laid the conceptual foundations of a discipline that formally took shape at the Dartmouth Conference in 1956, where the term artificial intelligence was first coined (Haenlein and Kaplan, 2019, 61). After years of limited progress and periods of stagnation (the so-called “AI winters”) the development of learning algorithms, particularly the technique known as backpropagation, in the 1980s made it possible to build systems capable of learning from data (machine learning). These advances, combined with the exponential growth of computing power and the massive availability of digital information, provided the foundation for today’s AI revolution.

In 1997, IBM’s Deep Blue defeated world chess champion Garry Kasparov. Although the system was not truly AI in the sense we understand it today (its strength relied primarily on brute computational power) the event symbolized the moment when a machine surpassed a human being in an activity long regarded as one of the highest expressions of rational intelligence. The relentless advance of Moore’s Law (the accomplished prediction that computing power would double approximately every two years) made it possible to build systems with increasingly dense networks of nodes. A critical turning point came in 2016, when AlphaGo, a neural network developed by DeepMind, a startup acquired by Google, defeated the world champion of go, South Korean player Lee Sedol. Unlike Deep Blue’s victory, AlphaGo’s success was not the result of computational force, but rather of a form of strategic instinct or intuition. AlphaGo prevailed with a move that Lee Sedol himself described as “sublime.” The achievement demonstrated that AI could generate surprising solutions or insights to complex situations that had previously been unknown to humans.

Conversational machines and their unexplainable nature

A new leap toward generative AI occurred with the introduction of the transformer architecture, through the influential paper Attention Is All You Need by researchers at Google (Vaswani et al, 2017, 30). The transformer represented a breakthrough AI architecture that enabled machines capable of developing and using natural language. From that point onward, OpenAI’s GPT (Generative Pre-trained Transformer) models emerged as the leading force in the field. ChatGPT, an artificial neural network with 175 billion parameters, became the first mass-market AI application. It was the fastest-adopted technology in history, reaching 100 million users in just two months. This phenomenon triggered a global race involving technology companies, startups, governments, and academic institutions. These systems came to be known as Large Language Models (LLMs). Their operation is based on artificial neural networks, with hundreds of billions, even trillions, of parameters (“neurons”). Their complexity is such that they are effectively “unexplainable”: even the engineers who designed them are unable to fully interpret the flow of information circulating within them. Yet the outcome is remarkable. These models exhibit “cognitive” capabilities that, across multiple dimensions, are comparable to those of humans. These capabilities include natural language generation, image recognition, computer programming, logical reasoning, mathematical problem-solving, or even the ability to answer PhD-level scientific questions.

Organizational implications: Toward the AI-Organization

Throughout history, every major technological revolution has given rise to a distinct form of organization. Firms adapt to their technological environment, seeking the most efficient structure for each era. Electricity made the assembly line and mass production possible. The internet gave birth to global digital platforms such as Google and Amazon. Mobile communications enabled the emergence of network-based companies such as Uber and Airbnb. In the same way, AI is creating the conditions for a new generation of organizations. When companies such as Google declare themselves to be “AI-First”, they are signaling what is likely to become the dominant organizational model of the coming decades. If the iPhone was the “internet phone” the challenge now is to imagine and design the “AI-phone” What will the AI-bank, the AI-supermarket, or the AI-McDonald’s look like? These organizations will not be built simply by incorporating AI into existing processes. Rather, entirely new organizational architectures will emerge, designed from the ground up around AI.

My hypothesis is that the AI-organization will be structured around three fundamental elements:

• The first will be a digital core: a system built around a vast data lake capable of compiling all the organization’s relevant information (internal parameters and key performance indicators, customer data, market trends, competitor movements, new patents and data on emerging startups, scientific publications, as well as macroeconomic indicators and trends captured from social media). In short, all information relevant to the company will be consolidated within this data lake.

• The second element will be an AI engine, most likely a Large Language Model (LLM) powered by large-scale cloud-based supercomputing infrastructures. This engine will function as a form of organizational intelligence, trained with the data of the digital core, capable of detecting patterns invisible to humans, identifying emerging opportunities, generating working hypotheses, designing new products, and anticipating threats.

• The third element will be a human decision layer. We must remember that AI systems operate as opaque, non-deterministic “black boxes”, with a stochastic component influencing their outputs and that these outputs are also conditioned by the training strategy and the input data. Therefore, contrary to some futuristic visions that foresee fully automated enterprises, I argue that the AI organization will continue to require strong human oversight, at least in the most critical activities – those that have a clear impact on the business performance and/or on the safety or dignity of human beings. This brings us to a longstanding principle in leadership: virtually every managerial function can be delegated, with the only exception of accountability. In a world of electronic minds, responsibility for the consequences of machines’ actions will remain human. Social, economic, and legal responsibility cannot be diluted within a dense network of artificial neurons, autonomous algorithms, and remote digital connections. AI will be able to analyze complex information, make recommendations, and propose solutions, but final responsibility will remain in human hands. This will reshape the focus of leadership: as AI systems increasingly participate in the interpretation of the business environment, managers will need to devote less time to processing information and more to exercising moral judgment (on the role of human judgment in management and entrepreneurship, see also Maggioni in this issue), framing strategic questions, and critically evaluating AI-generated recommendations.

Yet in an increasingly automated and interconnected world, it will also be necessary to orchestrate higher systems of decision-making. Social responsibility goes beyond individual accountability. Leaders will need to cultivate organizational communities capable of preserving trust, cooperation and shared purpose. A new interpretation of management must be developed in a global environment in which AI will be embedded in every piece of software. Magnifica Humanitas calls for “intermediary organizations and educational communities that rebuild trust and relationships“ (181). For this reason, we will need superior and new forms of governance, embedded in rules, regulations and institutions. As Magnifica Humanitas argues, AI governance requires forms of shared responsibility involving firms, educational institutions and civil society to assure “independent checks, transparency regarding algorithms, equitable access to data and avenues for recourse” (71) to ensure alignment with the common good. Corporate social responsibility (CSR) can therefore no longer be understood merely as a set of organizational policies or compliance mechanisms. In the AI age, CSR becomes an essential function of leadership itself: ensuring that technological capabilities are governed in ways aligned with the common good and preserving accountability for the consequences of organizational decisions.

Cognitive advantage, productivity and labor market

Just as economies of scale defined industrial competition and digital platforms created network effects, AI introduces a new source of competitive advantage: “cognitive advantage” (Ferràs-Hernández, 2026). Organizations equipped with AI will be able to learn faster, interpret reality more accurately, and transform vast amounts of data into actionable knowledge. This advantage will not stem from technology alone, but from the combination of superior computational capabilities and sound human judgment. Cognitive advantage will translate directly into higher organizational productivity through two complementary mechanisms. First, AI will augment human capabilities, enabling professionals to accomplish far more within the same amount of time. Second, organizations will increasingly substitute human labor with machines across a growing range of cognitive and operational tasks. Both mechanisms promise significant gains in efficiency and competitiveness. Unlike previous sources of competitive advantage, cognitive advantage is not merely a technological asset; it may reshape the distribution of economic value between labor and capital.

Yet while economic history suggests that disruptive technological revolutions have ultimately created more jobs than they destroyed, the unprecedented speed and breadth of AI make it uncertain whether this pattern will be repeated. As Magnifica Humanitas reminds us, “every real transition involves discontinuities” (153). The productivity gains generated by AI may therefore be accompanied by unemployment, widening inequality, and new forms of social exclusion if their benefits are unevenly distributed. Firms may be tempted to adapt human work to the logic and speed of intelligent machines rather than designing technology around human flourishing. The challenge for future leaders is consequently not only to maximize efficiency, but also to reconcile productivity with dignity and meaningful work. For this reason, the transition to an AI economy cannot rely on technological innovation alone. Recent contributions to the AI policy debate increasingly acknowledge that the unprecedented productivity gains enabled by artificial intelligence may be accompanied by profound social disruption. Magnifica Humanitas argues, “solutions must be sought at national and local levels through the involvement of intermediary communities” (153). Institutions as diverse as large AI companies, leading economists, and policymakers have begun advocating new forms of social innovation – including sovereign AI funds, AI dividends, adaptive welfare systems, and new ownership models – to ensure that the benefits of cognitive technologies are more broadly shared. This emerging consensus reinforces the central insight of Pope Leo XIV’s encyclical: technological progress and business competitiveness is insufficient unless accompanied by institutional innovation.

What leaders need to learn. Strategy, technology and humanism

The emergence of new forms of AI, such as generative AI, has overturned many of the assumptions we held only a few years ago regarding the limits of machines. We once believed that analytics, planning, and finance would be the areas most affected by AI substitution, while social interaction, creativity and strategy would remain firmly under human control (Ferràs-Hernández, 2018, 27). Today, however, we see AI advancing precisely into these higher-value domains. We will increasingly coexist with creative systems capable of dramatically expanding our ability to comprehend reality, and to propose new disruptive ideas or business models. Furthermore, AI can develop entire strategic plans at near-zero marginal cost. What is, then, management’s last stand? In this new context, I argue that executives will need to develop a new combination of skills. They must continue to cultivate strategic thinking and management knowledge. Leaders cannot afford to outsource their understanding of the fundamentals of management to machines. Foundational knowledge is the wellspring of the creativity and critical thinking required to evaluate or validate the decisions made by machines. Without a solid knowledge of basic concepts such as productivity or inflation, it is impossible to assess economic decisions. Without understanding the fundamental principles of positioning or market segmentation, AI-generated marketing strategies cannot be properly evaluated. New ideas cannot be assessed without first mastering the old ones.

At the same time, corporate leaders will require technological thinking, allowing them to grasp the potential of AI and set its limitations. Without understanding the nature of AI systems, it is impossible to fully understand their applications. Without recognizing that AI is a non-deterministic technology (subjected to stochastic variation) it cannot be deployed appropriately within organizations. Alfred Chandler, a renowned Harvard professor and a pioneer of modern management, articulated as early as 1962 one of the most influential principles in management theory: “structure follows strategy” (Chandler, 1962, 5). Structural decisions (whether concerning investment, hiring, or expansion) should be subordinate to strategy. Today, however, strategy itself is increasingly shaped by technology. Technology will cease to be merely a support function and will instead become a central element of business strategy. Now, strategy follows technology. Management disciplines, and management itself, will increasingly merge with computational sciences and technological knowledge. Corporate leaders must show a deep understanding of technology.

Yet the more powerful technology becomes, the more important human capabilities will be, to define a moral boundary for technological development. Leaders must be able to formulate the right questions, contextualize the answers provided by AI, and ultimately assume responsibility for decisions. Machines may diagnose, predict, and even imagine alternative futures, but it will remain the responsibility of leaders to decide which path to follow, which risks to accept, and which values to preserve. Leaders will need a strong humanistic foundation, enabling them to interpret AI-generated insights within complex ethical and social frameworks. Corporate leaders will require a profound grounding in philosophy and the humanities to exercise their responsibilities wisely.

Conclusion

AI will redefine the discipline and practice of management. As organizations increasingly derive competitive advantage from cognitive capabilities, the role of leaders will shift from processing information to exercising prudent judgment, ethical discernment, and responsibility. Contemporary management research suggests that AI will become embedded in virtually every organizational process. This perspective may enrich future reflection within Catholic Social Teaching by extending its ethical analysis from AI itself to the governance of AI-based organizations and AI-mediated leadership. In turn, Magnifica Humanitas offers a framework for ensuring that these transformations remain oriented toward human dignity and the common good. The future of firms, and the future of capitalism itself, will therefore depend not only on technological sophistication but also on their capacity to govern AI wisely, distribute its benefits fairly, and preserve the uniquely human responsibility for deciding what ought to be done. Only in this way can AI become not merely a driver of productivity and economic growth, but a genuine instrument of integral human development.

Acknowledgement

The author used ChatGPT as an assistance tool for translation and language editing. The ideas, arguments, analysis, and conclusions presented in this article are entirely original and the sole responsibility of the author.