The Digital Divide: Africa and South America Left Behind in the Global AI Race

The infrastructure for artificial intelligence (AI) is concentrated in only a few nations, resulting in an increase in digital inequality that impacts science, the economy, and geopolitical dependency, potentially exacerbating global disparities even further.

A recent study from the University of Oxford indicates that there are specialized AI data centers in only 32 countries. Over 150 nations lack the computational power necessary for training cutting-edge AI models. Researchers Wili Lehdonvirta, Zoe Jay Hawkins, and Boxi Wu identified the United States, China, and the European Union as the primary beneficiaries: American companies like Amazon, Google, and Microsoft operate 87 AI facilities globally, while Chinese suppliers manage 39, and Europe lags behind with only six centers.

The Oxford team systematically assessed the global distribution of AI infrastructure by analyzing the client websites of nine leading cloud providers. The findings are striking: **the U.S. holds a significant edge, while Africa and South America are nearly entirely excluded from the AI development landscape.** Most of the chips utilized in these data centers come from Nvidia—a U.S. firm whose graphics processors have become foundational to the AI boom.

The New York Times highlights this gap with real-life examples: OpenAI CEO Sam Altman recently visited a $60 billion project in Texas as part of the Stargate initiative, whereas computer science professor Nicolás Volovik from the University of Córdoba in Argentina conducts AI research in a repurposed classroom using outdated chips. In Kenya, startups like Qhala and Amini are hindered by a lack of local computing resources and must work overnight to rent overseas facilities at lower costs.

The United States and China leverage their technological superiority as a tool for geopolitical influence. Washington controls **global access to high-performance chips through export restrictions**, while Beijing offers state-backed loans to promote its own hardware. In the UAE, Chinese technology was excluded in exchange for **access to Nvidia and Microsoft products**.

According to The New York Times, companies like Huawei are attempting to upgrade existing data centers in Africa using Chinese chips. Though China still trails Nvidia technologically, Lasina Kone of Smart Africa views this as a practical solution, emphasizing Africa’s openness to collaborating with any supplier capable of providing graphics processors.

Research co-author Lehdonvirta cautions that countries with computing power may wield influence comparable to oil producers. The high cost of Nvidia’s graphics processors makes them difficult to acquire, leading many countries to rent computing resources from distant data centers. In response, several nations are currently investing billions to develop their own AI infrastructure.

Governments in India, Brazil, and the EU are funding local data centers and AI models. In Africa, Cassava Technologies plans to establish five data centers with support from Nvidia and Google.

However, according to Cassava’s own estimates, these initiatives will only meet **a small fraction of the continent’s needs.** Oxford researchers warn that without broader access to computing resources, this gap is likely to **widen.** As Kone puts it, computing resources are becoming a cornerstone of digital sovereignty.

You can get 100,000 free tokens for your first tasks on BotHub by following this [link](https://bothub.chat/?invitedBy=m_aGCkuyTgqllHCK0dUc7) and start working right away!

[Source](https://the-decoder.com/african-and-south-american-countries-are-almost-entirely-excluded-from-global-ai-development/)