Radio ExpressTV
Live
How Developing Countries Can Maximize the Benefits of Artificial Intelligence
Shameeka Siriman: Senior Advisor to the Secretary-General of the United Nations on Trade and Development (UNCTAD).
Tafari Tsfashio: Senior Advisor at the Tony Blair Institute for Global Change.
Since the global financial crisis erupted in 2008, industrial policy has resurfaced in respected economic discourse after decades of being dismissed as misguided interventionist policy, particularly when adopted by developing countries. Today, its resurgence is spearheaded by advanced economies that once rejected it, driven by the rapid shift towards the adoption of artificial intelligence and renewable energy.
This renewed emergence presents new opportunities for developing countries, provided they can overcome three main obstacles: a weak enabling environment (including inadequate infrastructure and other necessary inputs), limited autonomy in policymaking, and financial constraints. Industrial policy is often understood as subsidies and tax breaks, but many developing economies must resort to far more than that. Without reliable digital connectivity, dependable energy supplies, trustworthy data protection systems, and a skilled workforce, ambitions for AI-driven growth will remain confined to polished rhetoric.
Public policy options in developing countries are also constrained, as World Trade Organization (WTO) rules limit the use of tools—such as export-related subsidies, local content rules, and technology transfer requirements—that were once the foundation of manufacturing success, particularly in East Asia. In contrast, major economies continue to pursue their own paths, implementing large-scale industrial policies that often diverge from or violate the rules expected of others. The disparity is evident: of the more than 2,500 industrial policy measures in force globally in 2023, these three economies accounted for nearly half.
Lastly, financial constraints are exceedingly harsh. In many developing economies, up to 80% of public spending goes to wages and debt servicing, leaving little for the long-term investments needed for manufacturing. Unlike the United States or the European Union, poorer countries cannot afford to provide massive support packages or finance multi-billion-dollar technology programs. Although technology clusters and business incubators have proliferated in Africa and Asia, few have managed to achieve tangible results. As noted by the United Nations Conference on Trade and Development (UNCTAD), these regions only succeed when integrated into established supply chains. Absent this foundation, they risk becoming costly, illusory projects—impressive on paper but unproductive in practice.
A reasonable approach to harnessing artificial intelligence in developing countries is to deploy existing advanced models. Freed from outdated infrastructure burdens, developing nations can leap directly to emerging technologies, much like many did when they bypassed landlines for mobile phones. Implementing AI costs a fraction of the expenses associated with building it. Anyone can utilize tools like ChatGPT without the need for building data centers or assembling specialized engineering teams.
Such targeted applications could be transformative. In healthcare, AI-supported diagnostic tools can help optimize scarce clinical capacities. In education, digital platforms can address the chronic shortage of teachers. In agriculture, predictive analytics can assist farmers in navigating climate variabilities. While these uses may not dazzle those at the forefront of technology, they can yield real returns in areas of paramount importance.
These applications also represent industrial policy at its most effective—practical, experimental, and oriented toward local realities. As Dani Rodrik asserts, “Success does not come from adhering to a fixed plan but from identifying sectors where public measures can unleash their full potential.”
Certainly, even a modest innovation agenda requires funding, and local venture capital remains scarce in many developing economies, where private wealth often tends to migrate abroad. However, governments can establish institutions to mobilize larger amounts of private capital, such as blended finance, sovereign innovation funds, directed guarantees, and regional technology centers. Donors too can (and should) increase support. According to the Organisation for Economic Co-operation and Development (OECD), the information and communications technology sector barely receives 2% of total aid for trade payments, a far cry from what is needed to build digital capabilities.
Governments in developing countries must also leverage digital technologies to enhance efficiency, particularly in revenue collection, to create the much-needed fiscal space. UNCTAD’s work on digitizing customs, particularly through the Automated System for Customs Data (ASYCUDA), provides a useful illustrative example. In Angola, one of Africa’s largest oil-dependent economies, the shift to digital customs procedures yielded astonishing financial gains, with revenues rising by 44% in one year and 13% the following year as analogue barriers were dismantled.
In Iraq, the results were even greater. Once the main border points were digitized there, customs revenues soared by over 120% in a single year. In Bangladesh, one of the fastest industrializing economies in Asia, gradual digital reforms facilitated an average annual revenue growth of about 11% over several years, with improved compliance and a reduction in leakages.
While international cooperation remains essential, the rules of global trade must also evolve to better accommodate digital and green manufacturing strategies. The Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) of the WTO may have made sense in the era for which it was designed, but now it hinders access to vital technologies. Patent systems must empower wider dissemination, akin to the former compulsory licensing for life-saving medications.
Collaboration among developing countries is also crucial, as no single country can bear the scale of investment required for AI or clean technology alone. Joint platforms like CERN (for physics research) demonstrate how pooled expertise can help distribute costs, share risks, and unlock mutual benefits. A more promising approach lies in collective innovation. In much of the Global South—sharing similar pandemic burdens and climate risks, benefiting from an abundance of data, and relying on relatively low-cost technical talent—collaborative innovation is not only cost-effective but also a strategic necessity in an increasingly multipolar world.
The return of industrial policy marks a significant shift in global economic thinking, but for developing countries, it is both a blessing and a curse. The path to manufacturing has become more arduous, narrower, and constrained by stricter technological and regulatory standards. However, the challenge is not insurmountable. By investing in foundational capabilities, targeting high-impact artificial intelligence applications, mobilizing innovative financing, and leveraging available political space, countries can accelerate their development efforts. Success will not hinge on replicating the models of wealthy nations but on practical adaptation to local realities.
