This opinion draws on the arguments of Daron Acemoglu, Nobel Prize-winning economist and MIT professor, in his discussion on the World Economic Forum’s Radio Davos podcast
Artificial intelligence is not simply another technological innovation. It is becoming a force capable of reshaping economies, labour markets, political institutions and the relationship between citizens and the state. Its consequences will extend far beyond the countries and companies developing it. Yet a fundamental question remains largely unresolved: who gets to decide how this transformation takes place, and whose interests will guide it?
In a recent discussion on the World Economic Forum’s Radio Davos, Nobel Prize-winning economist and MIT professoArtificial Intelligence and the Crisis of Liberal Democracy: Who Gets to Shape Our Future?
r Daron Acemoglu raised an important democratic concern. The development of artificial intelligence is increasingly concentrated in a small number of powerful companies, particularly in the United States and China. Their decisions may influence the lives of billions of people, including those who have little or no representation in the debates shaping this technological future.
Acemoglu’s central argument is that technological transformation cannot be left exclusively to corporate executives or the interests of a handful of powerful institutions. Even when innovation is driven by private companies, its consequences are public. Decisions about how artificial intelligence is designed, deployed and integrated into economic life should therefore involve democratic participation. The future of society cannot become the private domain of those who control its most powerful technologies.
This concern is closely connected to a wider crisis: the weakening of liberal democracy and the erosion of the economic and social foundations that once sustained it.
The democratic bargain and its breakdown
Liberal democracy represents one of humanity’s major institutional achievements. It combines individual freedoms with political participation and public institutions that enable people to pursue their aspirations. Freedom of expression, association and religion are essential, but freedom also requires opportunities: access to education, infrastructure, economic security and protection against circumstances that prevent individuals from earning a livelihood.
Democracy, in this understanding, is not limited to holding elections. It is also about building institutions through which citizens can participate in public life, influence collective decisions and share in the benefits of economic progress.
Acemoglu argues that this achievement is now under pressure. Anti-establishment movements have gained strength, while public confidence in established democratic institutions has weakened. The danger is not simply political instability. It is the possibility that societies may turn towards forms of governance that restrict individual freedoms, weaken democratic participation, tolerate arbitrary authority and undermine equality before the law.
The roots of this crisis, in Acemoglu’s analysis, extend back approximately five decades, to the transition from industrial economies towards post-industrial systems shaped by digital technologies and globalisation.
During much of the industrial era, economic growth created a relatively direct relationship between business expansion and improvements in workers’ livelihoods. Businesses needed workers to increase production, while organised labour and strong trade unions helped workers negotiate better wages and working conditions. Greater commercial activity could therefore generate benefits for both businesses and employees.
This relationship was never perfect, but it helped sustain a broader social bargain: economic growth could translate into shared prosperity.
The emergence of digital technologies changed the terms of that bargain.
Technology, automation and the changing nature of work
Digital technologies have increasingly favoured workers with higher levels of education, particularly those able to perform managerial, administrative and data-related tasks. At the same time, automation has enabled businesses to replace tasks previously performed by human workers with machinery and algorithms.
The consequences extend beyond the loss of particular jobs. Automation can alter the distribution of economic power by reducing the demand for certain categories of labour while increasing the value of specialised skills and technological ownership. If productivity gains primarily benefit companies and highly skilled groups, economic growth may continue without producing comparable improvements in the livelihoods of a broad section of society.
The problem, therefore, is not technological progress itself. Digital technologies and artificial intelligence can create new opportunities, improve productivity and expand human capabilities. The more important question is how these technologies are developed and what economic purpose they serve.
Acemoglu’s argument invites a distinction between technology that complements workers and technology that primarily replaces them. A technological system designed around human capabilities can improve productivity while strengthening the contribution of workers. A system driven mainly by labour substitution may increase output while weakening employment opportunities and bargaining power.
This distinction should be central to public policy. Governments and businesses must consider not only how much technology can produce, but also how its benefits and costs are distributed across society.
Economic inequality becomes a democratic problem
The transformation of work has also contributed to a wider social divide. Acemoglu describes how workers, managers and highly educated professionals have increasingly become separated into different communities and social networks. The decline of shared workplaces and neighbourhoods has weakened some of the social connections that previously brought different economic groups together.
His discussion of the American industrial experience illustrates this change. He refers to a factory community in Ohio where workers, office employees and managers once lived in close proximity and sent their children to the same schools. Over time, these shared social spaces became less common. Economic and educational differences increasingly translated into separate neighbourhoods and communities.
This separation matters because democracy requires more than formal political institutions. It also depends on a degree of social cohesion and a shared understanding of common interests.
Acemoglu links this development to changes in political coalitions. Centre-left parties in several industrialised democracies historically drew substantial support from working-class voters and were associated with the expansion of welfare institutions, workers’ rights and policies aimed at shared prosperity. Over time, however, their support increasingly shifted towards highly educated populations in large urban centres, while sections of the working class moved towards centre-right and hard-right political movements.
This realignment has created a challenge for liberal democracy. When economic groups become socially separated and political parties lose their connection with sections of the population, democratic institutions struggle to represent a broad range of interests. Economic inequality then becomes a political and institutional problem, not merely a question of income distribution.
Artificial intelligence must serve society, not replace democratic choice
The debate over artificial intelligence brings these challenges into sharper focus. If the technology is developed primarily to automate human work, concentrate economic power and increase the influence of a small number of companies, it may deepen existing inequalities. If it is developed to complement workers, improve public services and expand access to knowledge, it could contribute to a more inclusive economy.
The difference will not be determined by technology alone. It will depend on institutional choices, public policy and the distribution of decision-making power.
This is why Acemoglu’s call for democratic input into the future of AI deserves serious attention. Six billion people outside the United States and China will also experience the consequences of technological decisions, yet they have limited influence over the institutions and companies shaping these developments. Their economic futures should not be treated as secondary considerations.
The responsibility of democratic governments is to establish rules that protect public interests, encourage innovation and ensure that technological progress does not become a mechanism for concentrating power. This requires public debate about employment, education, data, competition and the accountability of technology companies. It also requires institutions capable of representing citizens who may otherwise have little influence over technological decisions.
For developing countries, including Pakistan, this debate is particularly relevant. The future of work, the structure of employment and access to technological opportunities will be influenced by developments taking place far beyond national borders. Countries that do not participate meaningfully in shaping these debates risk becoming passive recipients of technologies designed according to the economic priorities of others.
Pakistan must therefore approach artificial intelligence as a question of economic policy, institutional capacity and democratic governance. The objective should not be to resist technological change, but to develop the capacity to understand it, regulate its consequences and ensure that its benefits reach beyond a narrow group of firms and highly skilled professionals.
That requires investment in education and skills, stronger public institutions, informed legislative debate and a policy framework that recognises workers as participants in technological progress rather than merely costs to be reduced. It also requires a public conversation about the relationship between innovation, employment and economic opportunity.
The larger lesson from Acemoglu’s analysis is that technological advancement does not automatically produce social progress. Nor does economic growth guarantee political stability or shared prosperity. These outcomes depend on the institutions through which economic and political power is organised.
Artificial intelligence may transform the way humanity works and lives, but the direction of that transformation remains a matter of public choice. A democratic society cannot surrender that choice entirely to corporations, just as it cannot allow economic growth to become detached from the welfare of its citizens.
The future of AI should be shaped not only by those who develop the technology, but also by the people whose lives it will transform. The central question is not simply how intelligent machines can become, but whether our institutions can remain democratic, inclusive and accountable as technology changes the world around us.
This opinion draws on the arguments of Daron Acemoglu, Nobel Prize-winning economist and MIT professor, in his discussion on the World Economic Forum’s Radio Davos podcast and his book, What Happened to Liberal Democracy.
Republic Policy | Governance. Institutions. Public Policy.
