In a stark illustration of the widening global divide in artificial intelligence (AI), Sam Altman, CEO of OpenAI, recently visited the construction site of a new data centre in Texas. Valued at an estimated $60 billion, this facility is projected to become one of the world’s most powerful computing hubs by next year. Meanwhile, in Argentina, Nicolas Wolovick, a computer science professor at the National University of Cordoba, operates one of the country’s few AI computing centres—housed in a converted room with outdated technology. “Everything is becoming more split,” Wolovick remarked. “We are losing.”
The disparity in AI capabilities underscores a growing digital divide, dividing nations with advanced computing resources from those without. This gap is reshaping geopolitical dynamics and global economies, creating dependencies as countries strive to avoid exclusion from a technology race that promises to redefine economies, enhance scientific advancement, and alter daily life.
The major beneficiaries of this technological evolution are the United States, China, and the European Union. According to research compiled by Oxford University, these regions host over half of the world’s most powerful data centres, crucial for developing sophisticated AI systems. Only 32 countries, or about 16 percent of nations, possess high-capacity computing facilities, often referred to as “compute power.” Notably, the U.S. and China dominate the market, operating more than 90 percent of the global data centres employed for AI development.
In stark contrast, regions such as Africa and South America lack significant AI infrastructures, while India has made strides with at least five centres and Japan with four, according to the Oxford data. More than 150 countries have no computing hubs at all.
The hurdles to establishing effective AI data centres are significant. These facilities are not only vast and energy-intensive but also require substantial investment and infrastructure that many countries struggle to provide. The concentration of ownership among a few tech giants exacerbates the situation, highlighting the immediate effects of this computing power imbalance.
AI systems, including widely used applications like OpenAI’s ChatGPT, are notably more efficient in English and Chinese—languages predominant in the regions where computing power is concentrated. As tech giants leverage AI for data processing, automation, and new service development, nations without robust computing resources face limitations in scientific achievements, business growth, and talent retention. Some officials have expressed concern about becoming overly reliant on foreign entities for essential computing resources.
According to Vili Lehdonvirta, a professor at Oxford University, “Oil-producing countries have had an oversized influence on international affairs; in an AI-powered near future, compute producers could have something similar, since they control access to a critical resource.” The strategic importance of AI computing power has made components like microchips integral to the foreign and trade policies of both the U.S. and China.
In response to the growing divide, some countries are investing public funds into AI infrastructure to enhance their technological sovereignty. The Oxford researchers mapped the world’s AI data centres, revealing that U.S. companies operate 87 AI computing hubs—approximately two-thirds of the global total. In contrast, Chinese firms manage 39, with European companies reporting only six.
“We have a computing divide at the heart of the AI revolution,” stated Lacina Kone, director general of Smart Africa. He emphasized that the issue extends beyond hardware availability; it involves the sovereignty of digital futures.
Historically, a technological gap has existed between wealthy and developing nations. Over the past decade, advancements such as affordable smartphones and increased internet access have led some experts to believe this gap has narrowed. As of last year, 68 percent of the global population accessed the internet, a significant increase from 33 percent in 2012, according to the International Telecommunication Union.
Despite these advancements, the UN cautioned in April that the digital gap may widen without proactive measures regarding AI. The report indicated that just 100 companies, predominantly from the U.S. and China, accounted for 40 percent of global investments in AI, cementing their control over the future of this transformative technology.
At the heart of the divide lies the demand for graphics processing units (GPUs), essential for AI development. These high-performance chips, primarily manufactured by Nvidia, require multibillion-dollar factories to produce. As demand surges, prices for these critical components have skyrocketed, leading to fierce competition among countries and companies for access.
While wealthier nations have the necessary infrastructure for AI, many others are left behind. Renting computing power from distant data centres poses challenges, including high costs, slower connection speeds, legal compliance issues, and vulnerability to shifts in foreign corporate strategies.
Brad Smith, President of Microsoft, pointed out that numerous countries seek greater computing infrastructure as an assertion of sovereignty. Yet, achieving this will be challenging, particularly in regions like Africa, where reliable electricity remains a significant obstacle. Microsoft is currently developing a data centre in Kenya in partnership with G42, a United Arab Emirates-based company, choosing locations based on market needs, energy availability, and skilled labour.
Alarmed by the concentration of AI capabilities, several countries are taking measures to bridge the gap. These initiatives include providing land and cheaper energy, expediting development permits, and investing public funds to acquire chips and build data centres. The aim is to establish “sovereign AI” that serves local businesses and institutions.
For instance, the Indian government is subsidising compute power and supporting the development of AI models tailored to the country’s diverse languages. In Africa, regional governments are collaborating on establishing compute hubs, while Brazil has committed $4 billion to AI projects. The European Union has outlined plans to invest 200 billion euros in AI initiatives, including new data centres.
Despite these efforts, closing the AI divide will likely necessitate collaboration with tech powerhouses in the U.S. and China. The race for AI supremacy is far from over, and the stakes are high for nations seeking to secure their place in this rapidly evolving landscape.
