The resilience of complex societies depends on the ability to transform economic change into social continuity. For more than a century, work has fulfilled this function as a silent stabilizer, absorbing technological shocks and redistributing over time the benefits and costs of innovation. Artificial intelligence now intervenes in this mechanism with a depth that goes beyond immediate employment impacts, placing under strain the link between work, life expectations, and collective cohesion. The issue does not concern the quantity of work available, but its capacity to sustain social order and institutional trust in contexts characterized by high economic complexity.

Signals of structural exclusion constitute the first indicator of this fragility. Analyses by the OECD and the World Economic Forum show that the widespread adoption of advanced automation systems does not produce uniform mass unemployment, but rather a persistent selection along lines of skills, adaptability, and access to cognitive infrastructures. According to the Future of Jobs Report 2023, more than 40 percent of the skills currently used in advanced economies will undergo significant transformation within five years, while the capacity to reabsorb lower-skilled workers remains uneven. Exclusion thus takes on a cumulative and silent form, made up of interrupted trajectories and a progressive loss of economic relevance.

This dynamic fuels redistributive tensions that go beyond the wage dimension. When a growing share of value is generated through technological capital and positional rents linked to the control of data and models, redistribution mediated by work loses effectiveness. Fiscal and welfare systems, designed on a relatively stable employment base, struggle to intercept these new sources of wealth. According to estimates by the International Monetary Fund, in high-income countries the share of labor income in GDP has been in steady decline for more than two decades, while components linked to capital and intangibles have grown. The redistributive conflict thus shifts from the terrain of wage bargaining to that of the legitimacy of contributory bases.

The social conflicts that emerge from this imbalance no longer follow the traditional lines of fracture between capital and labor. They manifest as tensions between those included in and those excluded from the circuits of innovation, between those who benefit directly from automation and those who suffer its indirect effects without accessing productivity gains. These are fragmented conflicts, often lacking organized representation, but potentially destabilizing because they are linked to the perception of economic superfluity. In the corporate world, this condition is recognizable in the growing distance between highly valued strategic functions and increasingly standardized operational areas, with effects on motivation and sense of belonging.

In this framework, Emile Durkheim’s reflection on anomie offers an interpretative key of particular relevance. Durkheim identified in the weakening of shared norms the principal risk to social cohesion in phases of rapid change. Artificial intelligence accelerates this risk not by eliminating rules, but by rendering them rapidly inadequate. Work continues to exist, but loses the function of symbolic and normative orientation that allowed individuals to situate themselves within a recognizable trajectory. The disconnection between individual effort and social recognition thus becomes a latent source of instability.

Companies occupy a central position in this transition. Strategies oriented exclusively toward the maximization of technological efficiency tend to underestimate the role of work as social infrastructure. Decisions that are rational in the short term can generate negative externalities that, over time, reflect back on the competitive environment itself. The growing polarization of skills, the reduction of intermediate positions, and the increase in professional uncertainty affect corporate reputation, the ability to attract human capital, and the social legitimacy of the firm. In markets characterized by high visibility and interdependence, these factors become structural elements of economic risk.

The scenarios that emerge oscillate between institutional adaptation and social rupture. Adaptation requires a profound rethinking of the relationship between work and stability, recognizing that work can no longer be the sole vehicle of redistribution and identity, but remains a central node around which to reorganize public policies and corporate strategies. Rupture manifests when these transformations are ignored, allowing tensions to accumulate until they are expressed in forms of conflict that are difficult to govern, as demonstrated by recent episodes of social polarization in highly automated economies.

Decision-making processes mediated by artificial intelligence further amplify what is at stake. Computational capacity reduces certain cognitive limits, but tends to treat social complexity as an external variable, difficult to model. Decisions appear optimal on a technical level, but may prove shortsighted with respect to long-term systemic effects. Conscious leadership is called upon to integrate into strategic choices a broader view of social consequences, not for abstract ethical reasons, but to preserve the sustainability of the economic ecosystem in which it operates.

Social stability in the era of artificial intelligence thus takes the form of a dynamic equilibrium, continually exposed to forces of disintegration and recomposition. Work remains the primary indicator of a society’s capacity to include, orient, and recognize its members, even when it loses quantitative centrality. In this space of transformation, technology does not automatically determine outcomes, but makes institutional and organizational responsibilities more visible. The resilience of complex societies will depend on the ability to construct a new relationship between value, work, and cohesion, accepting that economic innovation does not replace, but makes more urgent, a collective choice about the type of social order one intends to sustain.

Global AI Observatory