Artificial intelligence and advanced automation are affecting the functioning of capitalism more deeply than previous technological waves. This is not merely an increase in productivity or a change in industrial processes, but a transformation that affects the cognitive and organizational foundations of accumulation. The economic system continues to operate according to logics of concentration, rent extraction and control of strategic resources, but the resources themselves change in nature. Data, computational capacity and algorithmic infrastructures become central factors, redefining the relationship between value, labor and institutional power.
The continuities of capitalism remain evident. Capital maintains its structural tendency toward exclusive appropriation and the transformation of competitive advantage into a dominant position. According to estimates by the World Bank, more than 70 percent of the market value of large technology firms is today attributable to intangible assets, largely linked to proprietary software, data and models. This dynamic does not represent a break with the past, but its reworking. Artificial intelligence does not overturn the objective of accumulation, but refines its mechanisms, making control more pervasive and less visible.
The fracture emerges on the side of productive factors. For the first time in industrial history, productivity growth is not accompanied by a proportional growth in employment. OECD reports show that, in sectors with high automation intensity, the increase in value added is significantly higher than the growth of the labor force. Automation does not merely replace physical labor, but reduces the role of standardized cognitive labor, delegating to predictive systems decision making activities once entrusted to human experience and judgment. Labor remains central on the social and symbolic level, but becomes structurally less necessary as a direct input in the production of value.
This disconnection generates new economic conflicts. Tension no longer unfolds exclusively along the capital labor axis, but shifts toward the control of automation infrastructures. Those who govern computing platforms, models and data flows exercise a power that precedes the market and conditions its functioning. According to analyses by the International Monetary Fund, the concentration of computational power contributes to widening inequalities not only of income, but of decision making capacity. Access to forecasting and coordination becomes a differential resource, determining who can anticipate economic dynamics and who is forced to undergo them.
Firms that dominate technological ecosystems acquire a capacity for influence that extends along production chains and beyond sectoral boundaries. Observable examples in global digital markets show how algorithmic decisions related to visibility, pricing or resource allocation can rapidly redefine the competitive balance of entire sectors. Economic conflict thus shifts from the traditional site of production to the architecture of the systems that make it possible, transforming technology into a central terrain of power.
In this context, entrepreneurial leadership assumes a different meaning. Leading an organization increasingly implies the interpretation of signals produced by systems that learn from the past and suggest optimized actions. Decision making responsibility is distributed between humans and algorithms, creating a grey zone in which success is attributed to the model and failure falls on the organization. Studies by Harvard Business School indicate that many executives rely on algorithmic outputs even in the presence of qualitative uncertainties, due to the difficulty of justifying divergent choices in highly measured contexts. The causal chain between decision and outcome becomes opaque, with significant implications for corporate and institutional governance.
On a theoretical level, the capitalism of automation recalls Karl Polanyi’s reflections on the separation between economy and society. If in the twentieth century the risk was a disembedding of the market from social bonds, today a further fracture emerges between value production and human participation. The economy tends to become self referential, guided by metrics internal to algorithmic systems, while society struggles to recognize itself in the outcomes of processes it neither fully understands nor controls. This distance fuels a crisis of legitimation that is not immediately visible in macroeconomic data, but emerges in social and political tensions.
The evolutionary scenarios of capitalism remain open. One trajectory of adaptive continuity envisages the integration of automation within new fiscal and redistributive equilibria, capable of partially compensating for the loss of labor centrality. Another, more unstable trajectory sees the accentuation of fractures between the concentration of economic power and the institutional capacity to govern its effects. Between these extremes lie hybrid configurations, in which technological innovation and institutional fragility coexist, producing a dynamic but precarious equilibrium.
For firms, the issue does not take on an ideological character, but a strategic one. Understanding continuities makes it possible to avoid the illusion of a total break with the past; recognizing fractures allows anticipation of systemic risks and positioning opportunities. Automation can operate as a multiplier of value, but also as an accelerator of instability if it is not accompanied by reflection on the distribution of benefits and long term sustainability. In this space of tension between historical inertia and profound transformation, the conditions of legitimacy of contemporary capitalism are redefined, in a process that remains open and structurally unfinished.
Global AI Observatory