P B Mehta writes: As we contemplate possibilities of AI, it is wreaking enduring transformations in state-capital relations

There are two broad theories about the economic roots of the backlash against globalisation. The first, and familiar one, emphasises distributional conflict: Loss of manufacturing jobs, rising inequality, wage stagnation, and fears of cultural displacement. Globalisation produced aggregate gains, but it did not compensate its losers. The task, therefore, was political management and redistribution: Rebuild welfare states, cushion shocks and restore legitimacy. There is no fundamental transformation in state-capital relations; rather, politics failed to keep markets socially embedded. The second theory, more conspiratorial and likely a competitor rather than a complement, argues that the backlash reflects a structural transformation in capitalism itself. Technological change, culminating in artificial intelligence, has altered the interests of capital in ways that make the old globalisation model less attractive. The open liberal order has not primarily collapsed because of cultural resentment or even inequality. Those have always been longstanding features of capitalism. It is, rather, that AI requires reconfiguration of the strategic alignment between states and capital. The technological effects of AI, its implications for jobs, are much discussed. But what is palpable already is the way it is altering institutional possibilities globally. In the era of manufacturing, capital benefited from distribution across geographies. Firms arbitraged wage differentials, fragmented supply chains, and used the credible threat of exit to discipline both labour and governments. As Karl Polanyi once suggested, states were increasingly embedded in markets. Capital mobility constrained public power. Globalisation was not merely an ideology; it was grounded in a vision of production. It looks like AI may shift this calculation. First, as is obvious, frontier AI is extraordinarily capital-intensive and infrastructure-dependent. It requires massive computing power, advanced semiconductors, energy supply, and highly specialised talent pools. The firms leading this transformation — Apple, Microsoft, Amazon, and Google — no longer resemble “light” digital platforms. They look more like early 20th-century infrastructure monopolies: Vertically integrated, asset-heavy, and territorially anchored. Data centres, chip fabrication plants, and energy grids are not footloose assets. They bind firms to specific jurisdictions and regulatory regimes. Second, AI collapses traditional distinctions between market coordination and administrative control. In a Hayekian world, price was information. In an AI-driven world, information becomes price. Algorithmic systems can generate individualised pricing, predictive behavioural targeting, and real-time risk assessment. Markets become less spaces of decentralised discovery and more systems of computational allocation. But such systems depend on legal permissions, access to data, and national regulatory frameworks. Even when algorithms are privately owned, their operation at scale requires state sanction. The possibility emerges of a deeper symbiosis: Capital extracts data; the state gains surveillance capacity. Consumer service infrastructures and citizen surveillance architectures converge. The state helps capital, and capital helps the state control its citizens. Third, geopolitical competition has shifted from access to markets to control over technology. Industrial policy is less about reshoring old manufacturing jobs and more about commanding semiconductor supply chains, rare earth minerals, cloud infrastructure, and foundational models. The object of competition is technological sovereignty. Exit ceases to be an effective disciplining mechanism because the relevant assets, data centres, energy systems etc., are territorially embedded. This alignment is clearest in China, where the party-state has long fused political authority with strategic sectors. But the shift is not uniquely Chinese. The US, through export controls, semiconductor subsidies, and national security review of technology flows, is also consolidating a techno-nationalist framework. The language differs — innovation, competitiveness, resilience — but the structural logic is similar: Frontier technology requires state capacity, and states seek dominance through alignment with big capital. Big capital, in turn, is happy with helping states gain social control. In short, it is happy to align with authoritarianism. Two indirect pieces of evidence support this thesis. One, in populist movements, nationalism is less a redistributive programme and more an ideological veneer for elite state-capital coordination. Two, the libertarian utopianism that once animated the tech sector, cryptocurrencies beyond state control, spontaneous networked publics, the withering of hierarchy, has faded. Emerging, instead, is a techno-nationalist order comfortable with control, concentration, surveillance and strategic rivalry. This isn’t a re-embedding of markets in society, as many had hoped; it is a re-embedding of capital in the state. For countries like India and much of Europe, the challenge is distinctive. The issue is not simply regulatory architecture or redistributive correction. It is whether domestic capital and state capacity can combine at a sufficient scale to compete within this new techno-nationalist complex. The axis of global competition is no longer light versus heavy regulation, nor open versus closed markets. It is the degree to which states and large firms can fuse around infrastructure, data, energy, and defence. • 1If ‘miya-biwi razi’, then what is the role of parents in marriage? Gujarat must stop policing love • 2Did Youth Congress’s ‘shirtless protest’ at AI summit hurt India’s image? That is the wrong question • 3Devuji’s surrender marks the end of the Maoist movement in India • 4Evidence, not panic, must shape street dog policy • 5Rahul Gandhi’s self-righteousness will only harm him and his party’s prospects This also means that it is very likely that techno-nationalism, on its own, will not resolve the legitimacy crises facing society. It does not begin to address inequalities of power; indeed, it may well exacerbate them. If capital concentration increases, and with it the risks of financial instability, the potential for an even deeper backlash remains. And while this complex seeks to answer the problem of meaning by giving technology the grand veneer of saving national civilisation, the social ramifications of AI are far too uncertain to suggest that it can remotely address the crisis of meaning that supposedly animated the populist backlash. This interpretation certainly overstates the coherence of the transformation. Traditional institutions and markets will continue to exist; labour still matters; global interdependence has not vanished. Yet it is difficult to ignore the signs: The return of territorial nationalism, the erosion of the public-private distinction, the convergence of civilian and military technologies, and the declining relevance of exit as a source of discipline. The backlash, on this view, is less a populist revolt from below than a technology-driven realignment from above. We will soon find out which interpretation of the backlash against globalisation has more staying power. But even as we contemplate the possibilities of AI, let us not forget the transformation in state-capital relations it has already brought about. The writer is contributing editor, The Indian Express