BLOG POST

Three Reasons Why AI May Widen Global Inequality

Artificial intelligence (AI) is often hailed, especially by those seeking financing, as a transformative force capable of reshaping economies and societies. But the impact of AI on the global income distribution remains highly uncertain.

While AI will, hopefully, boost macro-level productivity, it could widen income disparities within countries, benefiting highly skilled workers, displacing lower-skilled jobs in repetitive tasks, and concentrating wealth among those who control the technology. But the bigger, and far-less explored, concern is the inequality AI could amplify between nations.

Catch-up growth has become stronger since the late 1980s and has led to absolute convergence between nations since 2000. But this trend is threatened by the recent erosion of democratic governance and the uneven impact of climate change.

This blog highlights a third risk: AI deployment. While AI may fuel within-country inequality, it could also slow or reverse the gains made in reducing between-country inequality, as the ability to harness AI’s benefits or mitigate its disruption varies widely across nations. Without targeted policy interventions, AI may deepen the global divide, advancing richer nations while leaving poorer ones further behind, hindering progress towards the Sustainable Development Goals.

Uneven prosperity, uneven disruption

We’ve seen it before: previous technological waves, like the Industrial Revolution and the rise of the ICT era, produced deep and persistent income inequality—benefiting wealthy nations much more than others. With AI, we may well find ourselves to be on a similar path for three reasons.

1. Richer countries are better equipped to harness AI’s benefits

High-income countries, along with wealthier developing nations, hold a distinct advantage in capturing economic value from AI thanks to superior digital infrastructure, abundant AI development resources, and advanced data systems (Figure 1).

Figure 1. AI Preparedness Index by Country Income Group, 2024

Line plot illustrating that high-income countries, along with wealthier developing nations, hold a distinct advantage in capturing economic value from AI

Source: IMF, 2024.

Note: Acronyms: high (HIC), upper-middle (UMIC), lower-middle (LMIC) and low (LIC) income countries. Countries with labels are those with a population over 25 million or those who are in the top or bottom five of each income group.

In 2023, the United States alone secured $67.2 billion in AI-related private investments, which was 8.7 times more than China, the second-highest country in this regard. This concentration of capital allows these nations to lead in AI innovation, with the US producing 61 notable AI models in 2023.

The gaps with most of the developing world are stark. Consider these statistics. China is the only country below the high-income threshold among the top 30 most innovative countries. Internet access is just 27 percent in low-income countries and 52 percent in lower-middle-income countries—compared to 80 percent and 93 percent in upper-middle and high-income nations. Fixed broadband costs account for just 1 percent of monthly GNI per capita in high-income countries, well below the UN Broadband Commission's affordability target of 2 percent. These costs climb to 3 percent in upper-middle-income countries, and 8 percent and 31 percent in lower-middle and low-income countries, respectively.

Given these preexisting disparities, two concerns emerge:

First, AI could reinforce the dominance of wealthier nations in high-value sectors like finance, pharmaceuticals, advance manufacturing, and defense. As richer countries use AI to enhance productivity and innovation, it becomes harder for poorer countries to penetrate these markets.

Second, while AI is poised to primarily disrupt skill-intensive jobs more prevalent in advanced economies, it can also undermine lower-cost labor in developing countries. Automation in manufacturing, logistics, and quality control would enable wealthier nations to produce goods more efficiently, reducing the need for low-wage foreign workers. This shift, supported by AI-driven predictive analytics and customization capabilities, may allow richer countries to outcompete on cost, speed, and product desirability.

In short, richer nations appear far better positioned to capitalize on AI’s benefits, potentially deepening existing inequalities.

2. Poorer countries may be less prepared to handle AI’s disruptions

Just as the benefits of AI may be unevenly distributed, so may be the disruptions it causes. While high-income countries may experience greater labor market displacement—given their larger share of skill-intensive jobs that are more susceptible to AI automation—they are much better positioned to manage these shifts. Their better-developed social safety nets and active labor market policies, such as in Germany, can cushion the blow, retrain displaced workers, and stimulate new job creation.

Conversely, poorer nations face a different reality. Limited resources and underdeveloped social protection systems mean they are less equipped to absorb the economic and social shocks caused by AI-driven disruptions. Many lower-income countries already struggle with high rates of informal employment and fragile labor markets, leaving workers highly vulnerable to sudden economic shifts.

The lack of fiscal space also restricts these countries from investing in crucial areas like reskilling programs, infrastructure upgrades, or targeted welfare schemes to support affected communities. Without such mechanisms, the impact of AI-related job losses could exacerbate unemployment and deepen poverty.

Poorer nations risk falling further behind as they lack the tools to effectively manage AI’s disruptive effects on their labor markets and economies.

3. AI is intensifying pressure on traditional development models

AI is challenging the development models that have driven growth in many emerging economies for decades. Historically, export-oriented manufacturing absorbed large numbers of workers—often shifting them from agriculture—and drove productivity gains as production processes became more sophisticated.

However, this model is now under pressure. Manufacturing is becoming more technology- and capital-intensive, reducing its ability to provide widespread employment. As a result, its share of GDP and employment in many developing countries is steadily declining.

Take Bangladesh, for example, where AI and robotics are already being integrated into various stages of the garment manufacturing process, from automated sewing to fabric inspection and cutting. While these technologies boost efficiency, reduce costs, and improve quality, they also threaten to displace a large portion of the workforce. Estimates suggest that as many as 60 percent of jobs in Bangladesh’s garment sectors could be lost due to automation by 2030.

What alternatives exist to manufacturing? While manufacturing should not be dismissed (as the sector is large and heterogeneous with plenty of opportunities abounding), focus is increasingly shifting to export-oriented services. Countries like the Philippines and India have seen success in business process outsourcing, thanks to booming call center industries and IT services. But AI poses a threat to this model as well. AI has the potential to reduce the labor intensity of these activities, eroding the competitive edge in the international marketplace of lower-cost service providers.

If AI were to undermine labor-intensive service industries, developing countries may find it harder to identify viable pathways for growth, posing a significant challenge to long-term development and dampening the prospects of convergence.

Policy priorities for the AI era

So, what can developing countries do to navigate these challenges and mitigate the potential negative outcomes?

Invest heavily in digital infrastructure. Robust internet and data systems are foundational for AI development and deployment. Without these, countries will find it difficult to participate in the AI-driven global economy. The internet should be considered as a public good, with public policy focusing on strengthening its availability, affordability, reliability, and inclusivity. A good example is India, where initiatives like "Digital India" have connected over 600,000 villages to high-speed broadband by 2023, laying the groundwork for AI’s growth in sectors such as education and agriculture.

Ensure reliable electricity. The use of generative AI is expected to drive a 160 percent increase in global power demand from the data centers that process that traffic by 2030. Strengthening energy infrastructure, for example by modernizing power grids, optimizing energy efficiency, and diversifying energy sources can help developing countries address the challenge of unreliable electricity. Take Kenya as an example, which managed to diversify its energy sources over the past decade and now generates over 40 percent of its electricity from geothermal power.

Emphasize education and skills training. Improving STEM education and vocational training will prepare the workforce for the jobs of the future, ensuring they are not left behind as AI transforms industries. Governments and educational institutions should work together to create curriculums that include AI, machine learning, and data science. Viet Nam, for example, has integrated AI and digital literacy into its national education strategy. By 2025, the country aims to include AI in all levels of education.

Foster international collaboration. Partnering with technologically advanced nations can help developing countries leapfrog some stages of technological development. These partnerships can also open opportunities for joint ventures and collaborations that can drive local innovation. Collaboration on commitments and actions outlined in the Global Digital Compact would ensure countries’ agency and inclusivity in global AI frameworks under the guidance of the UN.

Promote local innovation. Encouraging homegrown AI solutions that address specific regional challenges can lead to more effective and sustainable development outcomes. Local entrepreneurs and innovators should be supported through funding, training, and policy frameworks that enable experimentation and growth. Rwanda has promoted local innovation by launching coding boot camps and AI development programs, particularly for youth and women, to encourage grassroots digital innovation.

Prepare for potential job displacement. Regulatory measures should be put in place to address potential job displacement, ensuring that social safety nets and retraining programs are available for affected workers, as with Singapore’s SkillsFuture programs. Developing countries should also consider new fiscal policies that adapt to the evolving economic landscape. As labor’s share in production declines, government revenue will be significantly impacted unless new measures of taxing capital are devised.

With these policies, developing countries can better position themselves to harness the benefits of AI while mitigating its risks. This transformation will require collaboration across governments, the private sector, and international organizations to build a resilient, AI-enabled future.

Will global inequality rise or fall?

The rise of AI could exacerbate both within-country and between-country inequality, thus placing upward pressure on global inequality. High-income individuals and regions stand to benefit disproportionately, while lower-skill workers and resource-poor regions risk being left behind. Nations with advanced AI capabilities are pulling further ahead, deepening the global divide. However, the future is not set in stone.

Whether AI intensifies or reduces global inequality will depend on the policy choices we make today. Governments and international organizations must act swiftly to counter AI’s disruptive effects. Investments in education, reskilling programs, and social safety nets are essential to prevent widening disparities. Progressive taxation and wealth redistribution could help reduce inequality within nations, while international cooperation and targeted aid could support AI adoption in developing countries.

However, AI also holds the potential to reduce inequality—if harnessed for social good. AI-driven innovations in healthcare, education, and agriculture can uplift living standards in developing countries, closing the gap between rich and poor. For instance, precision farming powered by AI can boost crop yields for smallholder farmers, while AI-enabled education tools can bring quality learning to remote areas. Whether these advancements will lead to macro-level shifts remains to be seen.

Ultimately, ensuring that AI development is inclusive and benefits all requires global collaboration, ethical AI practices, and a commitment to leveraging AI for public services and underserved communities.

This blog is part of Project Vision—a new initiative by UNDP to examine how countries start and sustain future-focused reforms in this era of unmet aspirations, rising insecurity and heightened turbulence. It also aligns with our forthcoming flagship report, AI for Human Development (AI4HD). We gratefully acknowledge valuable feedback and contributions from Christophe Bahuet, Viola di Canossa, Sudyumna Dahal, Devika Iyer, Amos Peters, Shahid Yusuf, and Zhu Linghui.

Disclaimer

CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.


Image credit for social media/web: Ariel D. Javellana (ADB/2009) via Flickr