Artificial intelligence is not only a technological revolution. It is a political economy revolution. It is reshaping productivity, labour markets, institutions, competition, education, public finance and the distribution of opportunity.

Like electricity, the steam engine and the internet before it, AI is a general-purpose technology. But unlike previous waves of transformation, it is moving faster, reaching deeper into both cognitive and manual work, and concentrating value around data, computing power, talent and global platforms.

The key question is therefore not whether AI will change our economies. It already is. The real question is whether societies will govern this transformation in a way that expands prosperity, strengthens democracy and protects human dignity — or whether it will deepen inequality, concentrate power and weaken public trust.

The political economy of AI begins with a paradox.

On one side, AI offers one of the greatest productivity opportunities of our generation. It can accelerate research, improve public services, reduce administrative burdens, support medical diagnosis, optimise logistics, modernise education and help firms become more competitive. Stanford’s AI Index Report 2025 shows that AI adoption and investment are accelerating rapidly: global private investment in generative AI reached USD 33.9 billion in 2024, while 78% of organisations reported using AI, compared with 55% one year earlier (Stanford HAI, 2025).

On the other side, the same technology may disrupt labour markets, shift income from labour to capital, increase the premium for high-skilled workers and strengthen the position of already dominant firms. The IMF estimates that around 40% of global employment is exposed to AI, with exposure rising to about 60% in advanced economies (IMF, 2024). The ILO similarly emphasises that generative AI is likely to transform tasks within occupations, with different effects across countries, sectors and worker groups (ILO, 2023; ILO, 2025).

This is why AI must be understood not only as innovation, but as distribution. Not only as efficiency, but as power. Not only as digital transformation, but as institutional transformation.

The new productivity frontier

For economists, productivity is the foundation of long-term prosperity. Without productivity growth, wages stagnate, fiscal space narrows, social systems become harder to finance and convergence becomes slower. For small open economies, including those in the Western Balkans, the strategic question is whether AI can become a shortcut to productivity convergence.

AI can reduce the productivity gap. AI can reduce the distance between small firms and global knowledge. A small company can now access analytical tools, translation, design, coding support, market intelligence and administrative automation that were previously available mainly to large corporations. A public administration can use AI to simplify procedures, detect fraud, improve tax compliance, accelerate permits and provide better services to citizens. Universities can use AI to personalise learning and support research. Hospitals can use AI to improve early diagnosis and resource allocation.

But productivity gains do not happen automatically. History shows that general-purpose technologies require complementary investment: infrastructure, skills, organisational change, regulation, institutions and trust. Electricity did not transform factories simply because it existed. Firms had to redesign production. The internet did not transform economies simply because cables were installed. Business models, logistics, finance and human capital had to adapt.

AI will be the same. The countries that benefit most will not necessarily be those that first consume AI tools. They will be those that reorganise education, firms, public administration and regulation around the productive use of AI. The World Bank’s work on digital progress stresses that digitalisation remains uneven and that poorer economies risk falling further behind if they lack infrastructure, skills and institutional readiness (World Bank, 2024). Its 2025 report on AI foundations similarly highlights both the opportunities of AI and the challenges low- and middle-income countries face in adapting and deploying it effectively at scale (World Bank, 2025).

Labour, inequality and the future of the middle class

The political economy of AI will be judged by its impact on work, income distribution and power. Work is not only income. It is identity, dignity, social mobility and citizenship. If AI increases productivity but weakens the bargaining power of workers, prosperity will become politically fragile. If it creates wealth but concentrates ownership, societies will face a legitimacy crisis.

The IMF has warned that AI can affect inequality through employment, wages, productivity and capital income (IMF, 2024). The ILO’s refined global index of occupational exposure to generative AI shows that the impact of AI should be assessed at the task level, not only at the level of whole occupations (ILO, 2025). This distinction matters. AI may not simply “replace jobs”; it may reorganise them, automate some tasks, augment others and change the skills required to remain productive.

The most vulnerable workers may not only be those with low education. AI also affects clerical, analytical, legal, financial, programming, media and administrative tasks. Many entry-level roles may be restructured, which creates a particular risk for young people trying to enter the labour market. If the first step of the career ladder disappears, social mobility becomes harder.

The World Economic Forum’s Future of Jobs Report 2025, based on the perspectives of more than 1,000 employers representing over 14 million workers, identifies technological change, AI adoption, demographic shifts, geoeconomic fragmentation and the green transition as major forces reshaping jobs and skills between 2025 and 2030 (World Economic Forum, 2025). This reinforces the need for a new labour-market strategy: not passive protection, but active empowerment.

The policy response cannot be nostalgia. We cannot protect every old task. But we must protect people’s capacity to adapt. This requires investment in digital skills, critical thinking, creativity, entrepreneurship, vocational training and active labour-market programmes. It also requires stronger links between universities, companies and public institutions.

The objective should not be “humans versus machines.” The objective should be humans with machines — workers who use AI to become more productive, firms that use AI to become more competitive, and governments that use AI to become more effective.

The state in the age of intelligent systems

AI will also redefine the role of the state. Governments are under pressure to do more with limited resources: deliver better services, regulate complex markets, protect vulnerable citizens, manage crises and support competitiveness. AI can strengthen state capacity — but only if deployed responsibly.

AI can help governments automate services, improve decision-making, detect fraud and support public-sector productivity. But it also creates risks: biased data, weak transparency, privacy violations and overreliance on automated systems. Public decisions require legitimacy. Citizens must know when AI is used, how decisions are made, how errors can be corrected and who is accountable.

Efficiency without accountability is not reform. It is technocracy.

This is why governance matters. The OECD AI Principles promote trustworthy AI that respects human rights and democratic values and were updated in 2024 to reflect technological developments (OECD, 2024). For countries aligned with the European integration process, the EU AI Act provides a key governance benchmark. The European Commission presents the Act as a framework to ensure that Europeans can trust AI while addressing risks from certain AI systems (European Commission, 2024).

For the Western Balkans and other EU-oriented economies, AI governance will increasingly become part of institutional quality, rule of law, public administration reform, data protection, competition policy and digital market alignment. The European path will require not only adopting technology, but also adopting standards of trust.

A capable AI state should have four pillars: digital infrastructure, data governance, human capital in the civil service and ethical regulation. Without these, AI risks becoming another imported tool rather than a national development capability.

Competition, sovereignty and power

AI is also a question of market structure. The new economy is being shaped by those who control data, cloud infrastructure, chips, foundation models, platforms and talent. This creates risks of concentration. If a small number of global firms dominate the AI ecosystem, countries and businesses may become dependent users rather than active participants.

For small economies, the answer should not be technological isolation. That would be unrealistic and costly. The answer should be smart integration: building local capacity, regional cooperation, public-private partnerships, open data ecosystems, AI-ready education and targeted niches where domestic firms can compete.

The Western Balkans may not build frontier foundation models at global scale. But the region can develop applied AI in public administration, health services, agriculture, logistics, finance, tourism, manufacturing, language technologies and compliance systems. It can also become a space for ethical, multilingual and inclusive digital governance.

A new social contract for the AI economy

The political economy of AI requires a new social contract. The UNDP’s Human Development Report 2025 frames the age of AI as “a matter of choice”, arguing that AI’s impact on human development depends on the choices societies make today (UNDP, 2025). This is the right framing. AI is not destiny. It is a tool whose social value depends on institutions, incentives and values.

This new social contract should rest on five commitments (MUSTs).

First, productivity must be shared. If AI raises output, the gains should support higher wages, better services, stronger social protection and more investment in people.

Second, education must become adaptive. Schools and universities should not only teach students to use AI tools, but to think beyond them: to reason, question, create, verify and make ethical judgments.

Third, firms must be supported to adopt AI responsibly. Small and medium-sized enterprises need access to training, finance, advisory support and digital infrastructure. Otherwise, AI will widen the gap between large and small firms.

Fourth, governments must modernise themselves. Public institutions should use AI to improve service delivery, fiscal efficiency and policy design, while preserving transparency, accountability and human oversight.

Fifth, societies must protect human dignity. AI should never reduce people to data points. Decisions affecting rights, access to services, employment, credit or justice must remain explainable, contestable and accountable.

Conclusion: AI as a choice, not destiny

AI is often described as inevitable. Technological progress may be inevitable, but its social outcome is not. The distribution of gains, the protection of workers, the openness of markets, the integrity of information and the quality of institutions are all matters of policy choice.

For policymakers, the challenge is to move faster than disruption without becoming reckless; to regulate without suffocating innovation; to promote competitiveness without accepting inequality as collateral damage; and to build digital states without weakening democratic accountability.

The countries that succeed in the age of AI will not be those that simply automate the past. They will be those that redesign the future. They will invest in people, build trustworthy institutions, support innovation, protect fairness and understand that prosperity is not measured only by technological capability, but by the quality of life it creates.

The political economy of AI is therefore a test of leadership. It asks whether we can transform intelligence into inclusion, productivity into shared prosperity, and innovation into a stronger social contract. The future will not be written by algorithms alone. It will be written by the values, institutions and choices of the societies that govern them.

References

European Commission. (2024). AI Act: Regulatory framework for artificial intelligence. European Commission.

International Labour Organization. (2023). Generative AI and jobs: A global analysis of potential effects on job quantity and quality. International Labour Office.

International Labour Organization. (2025). Generative AI and jobs: A refined global index of occupational exposure. International Labour Office.

International Monetary Fund. (2024). Gen-AI: Artificial intelligence and the future of work. IMF Staff Discussion Note.

OECD. (2024). OECD AI Principles. Organisation for Economic Co-operation and Development.

Stanford Institute for Human-Centered Artificial Intelligence. (2025). AI Index Report 2025. Stanford University.

United Nations Development Programme. (2025). Human Development Report 2025: A matter of choice — People and possibilities in the age of AI. UNDP.

World Bank. (2024). Digital progress and trends report 2023. World Bank.

World Bank. (2025). Digital progress and trends report 2025: Strengthening AI foundations. World Bank.

World Economic Forum. (2025). The Future of Jobs Report 2025. World Economic Forum.

Linkedin:  https://www.linkedin.com/pulse/political-economy-ai-productivity-power-new-social-contract-besimi-ewk2f/