AI Hustle: Human Capital, Sovereignty, and the Global South

At a Glance

• While the United States and China compete for the world's top AI researchers, countries from the Global North and Global South take varied approaches to human capital in their quest for AI sovereignty.
• The middle powers of the Global South occupy a distinct lane in global efforts to achieve AI sovereignty as they balance strategic autonomy and economic development. Human capital is among their most important assets, and two issues are key: (1) converting "brain drain" into "brain circulation" and (2) upskilling their labor force while protecting against exploitation.
• Out-migration of AI talent is not necessarily a loss. It can offer gains for Global South middle powers who actively engage their diasporas to foster cross-border knowledge exchange and investment.
• Not "losing" in the race for AI sovereignty means, in part, avoiding exploitation while upskilling workers as labor markets and economies evolve in response to AI. For example, microwork provides an entry point into the global AI value chain, but can be exploitative and is vulnerable to automation, making it a precarious foundation for AI-focused development.

I. Digital Feudalism in the AI Age

In February 2025, over 100 countries gathered in Paris for the most closely watched AI event of the year. The AI Action Summit, the third in a series of annual convenings that began at Bletchley Park in 2023, was framed as a discussion on trust, safety, and innovation. But the real subtext of the summit was power. In his keynote address at the Grand Palais, French President Emmanuel Macron offered a warning about "a system where there are those who are in power-and the rest are vassals."

In today's digital feudalism, a handful of platform companies extract algorithmic rents, controlling access to services that have become indispensable to twenty-first century societies and economies. Data and compute, rather than tithes and forced labor, now shape hierarchies of power. Macron seeks AI sovereignty through a "third way" that counterbalances the United States and China by bringing together an international network of countries from the Global North and Global South, including India, which co-chaired the AI Action Summit with France.

This brief explores the challenge of AI sovereignty for a group of countries that occupy a unique position in the international order: the middle powers of the Global South, who view AI through the lens of development and the long shadow of colonialism. Fiscal constraints limit their ability to build competitive AI stacks, but as middle powers-with influential emerging markets and leadership roles in international bodies-they have leverage to shape global AI governance and power to assert agency. Above all, they have one resource in abundance: human capital.

In Section II, I locate middle powers of the Global South within the broader spectrum of strategic orientations toward AI, juxtaposing them with superpowers, middle powers of the Global North, and Global South countries that don't rise to middle power status. Section III examines the two most pressing dilemmas Global South middle powers face in stewarding human capital for AI sovereignty: brain drain and microwork. Section IV looks to new paths to meet these challenges.

II. The Digital Great Game: Who's Playing to Win?

Scholars define AI sovereignty as a policy approach aimed at ensuring that the development, control, and deployment of AI systems remain in the hands of domestic actors. States pursue AI sovereignty for many reasons, but one of the most critical is autonomy-avoiding dependence on foreign powers that could exercise undue leverage over domestic technological ecosystems.

This concern is especially relevant given the AI dominance of the United States and China, which command the most formidable compute resources, the most advanced foundation models, and the highest flows of private investment. Though China speaks the language of cyber-sovereignty, for China, as for the United States, the aspiration is supremacy: Both are playing to "win" the AI race. For most other countries, the goal is simply to be in the race at all.

For middle powers of the Global North-such as France, the U.K., Canada, South Korea, and Japan-being in the race means operating near the AI innovation frontier. These countries craft national AI strategies; they show up on international leaderboards; and they invest in talent and infrastructure. During the AI Action Summit, for example, France announced a $110 billion investment package to build out high-end computing power. By contrast, most of the African continent-excluding South Africa-is a "compute desert," lacking even the most basic local AI infrastructure. For many in the Global South, AI sovereignty may be a useful rhetorical frame, but it is not a policy goal that is achievable in the near term.

Between these two poles are the middle powers of the Global South-countries such as India, Indonesia, Brazil, Turkey, and South Africa. Within the global race to develop, deploy, and regulate AI, they occupy a distinct lane. As part of the Global South, they approach AI as a tool for development that promises to accelerate the Sustainable Development Goals (SDGs). Yet, as middle powers, they also hold significant leverage. Wielding more power than at any time since World War II, these "swing states" in the international system benefit from multipolarity and regional fragmentation. They have powerful markets that tech companies covet. They hold leadership roles in important bodies of global governance. In asserting nonalignment, they are already beginning to shape an international logic of the future.

They also possess demographic advantages. As OECD countries rapidly age and China confronts the prospect of demographic collapse, their populations are youthful and growing. Add in educational institutions that are competitive regionally and globally, and what emerges is one of their most important resources when it comes to AI sovereignty: human capital.

III. Talent Wars: The Global South's Dilemma

How to train, retain, and attract STEM talent; how to upskill and reskill domestic populations; how to accelerate digital literacy; how to manage the labor market disruptions that AI will inevitably summon-all countries, no matter their strategic orientations toward AI, are urgently grappling with these questions. For middle powers of the Global South, however, two challenges are distinctive: leveraging their diaspora of AI talent and avoiding exploitation in the AI food chain.

From Exodus to Opportunity: Diaspora Networks

Policymakers traditionally take a dim view of skilled workers migrating abroad. The sending countries-often already low on human capital-invest in training skilled workers but may fail to benefit from their contributions, while receiving countries gain a one-sided advantage. This is particularly acute with AI professionals, who are 80 percent more likely to migrate than other STEM workers, attracted in part by exorbitant salaries in the Global North.

More recent research complicates this view, however, highlighting increased trade between sending and receiving countries as well as "brain circulation," where skilled migrants return to their country of origin equipped with skills of even greater value.

Take India. Its prestigious network of engineering schools-the Indian Institutes of Technology (IITs)-produces world-class STEM talent. Much of that talent leaves the country. Of the top 100 scorers on the 2010 IIT entrance exam, 62 percent left within eight years, many to the United States, where Indians make up seven out of ten H-1B visa holders. But those who leave often thrive in a way that benefits their birth country. Indian-born entrepreneurs founded 11 percent of the top 50 U.S. AI startups-the highest share of any immigrant group-creating cross-border linkages that strengthen India's AI ecosystem, now the world's fourth strongest. While India has improved retention of AI talent since 2019, policymakers also understand that skilled migration is not necessarily a loss, but potentially a "win-win" for both India and destination countries.

For other middle powers of the Global South, India offers an important lesson. Training and retaining AI talent is crucial, but eliminating skilled migration is neither feasible nor desirable. Instead, countries should mobilize their diaspora communities for developmental goals; seek to convert brain drain into brain circulation; and view high-skilled migration corridors to the United States and other Global North countries as a strategic asset in the pursuit of AI sovereignty.

Digital Serfdom or Stepping Stone?

Microwork is the second distinctive challenge that middle powers of the Global South face in leveraging human capital for AI sovereignty. Often performed on smartphones, microwork involves tasks such as image tagging, text annotation, and data classification-critical for training and testing AI models. According to a 2024 World Bank report, 156 million people engage in microwork worldwide, most of whom are under 30 and hold a bachelor's degree. Some 45 percent of microworkers are women, a higher share of female labor force participation than the economy-wide rates in many low- and middle-income countries where this work is concentrated. As AI reshapes the global economy, microwork has emerged as a distinct form of digital labor. But for Global South countries, and especially the middle powers among them, does microwork represent a pathway to AI sovereignty by upskilling their workers-or a form of techno-feudalism?

Nowhere is this dilemma more acutely felt than in Kenya, a key player in the Global South's microwork economy. President William Ruto has championed Kenya as a "Silicon Savannah," positioning the country as a hub for foreign tech investment. The strategy has yielded results. In 2023, Kenya attracted the most start-up capital in Africa, outperforming even South Africa and Nigeria. But Kenya's efforts have come at a cost. A year earlier, a TIME investigation exposed the harsh treatment of Kenyan content moderation workers working for Meta through an outsourcing firm. In one of several lawsuits brought against Meta, workers alleged forced labor and abusive conditions, citing exposure to traumatic video material without adequate mental health support. Concerned about scaring off foreign investors, Ruto advanced legislation to shield tech firms from similar liability.

Not all microwork involves such traumatic material, but even in more benign forms, the industry presents a conundrum. On the one hand, microwork provides jobs to unemployed and educated youth, often at wages that are highly competitive by local standards; even in Kenya, the $1.46 per hour paid to Facebook content moderators tripled the minimum wage. On the other hand, microwork's gig-based nature leaves workers insecure and at the mercy of foreign tech companies. More importantly, microwork may be an industry with an expiration date, as AI increasingly automates the tasks that feed into its development. Moreover, the $15 billion that microwork will generate globally by 2030 pales in comparison to the estimated $20 trillion contribution of the AI systems that microworkers will help to train. Microwork may offer an attractive foothold to the global AI value chain, but the key is to use it as a stepping stone and move on quickly.

IV. Beyond Vassalage: Charting New Paths

For middle powers of the Global South, the stakes of today's AI gold rush are not just about accelerating the SDGs or maintaining strategic autonomy from more powerful actors. At stake, too, are the lines of inequality that divide the Global North and South. Will AI deepen global disparities, as some predict, creating a world of digital lords and vassals? Or could technology serve as a force for democratization rather than domination?

The promising approaches emerging from Global South middle powers suggest a more optimistic path is possible. By strategically engaging their talent diaspora, these nations can transform brain drain into valuable knowledge networks that strengthen their position in the global AI ecosystem. Similarly, while being cautious about exploitation in areas like microwork, they can use these entry points to build digital skills and technical capacity that serve longer-term development goals.

Ultimately, achieving true AI sovereignty for middle powers in the Global South will require balancing multiple priorities: developing domestic talent while leveraging diaspora networks; participating in global value chains while avoiding digital vassalage; and creating policies that promote both innovation and protection. As these nations continue to assert their agency in international forums and shape global AI governance frameworks, they are pioneering pathways to technological empowerment that could redefine what sovereignty means in the digital age.

This piece is part of our Who Controls AI?: Global Voices on Digital Sovereignty in an Unequal World collection. View the rest of the series.