A senior United Nations official is raising alarm about an emerging technological divide that could entrench global inequality for decades. The concern centers not on basic internet access, but on who gets to build, deploy, and benefit from artificial intelligence systems.
According to AI Weekly, Tomas Lamanauskas, deputy secretary-general of the International Telecommunication Union, has framed the problem as an "opportunity divide" distinct from the traditional digital gap. While roughly 2.2 billion people lack internet access entirely, Lamanauskas argues that the real long-term threat lies in the geographic concentration of AI infrastructure and development capability.
The Infrastructure Imbalance
The numbers underlying this concern are stark. Africa and Latin America combined control just 3.3 percent of the world's data center capacity, the computational backbone required to train and operate modern AI systems. This disparity matters because data centers represent not merely computing power, but economic opportunity. Nations hosting these facilities gain jobs, tax revenue, and control over how AI systems are designed and deployed within their borders.
The challenge extends beyond existing infrastructure gaps. As AI systems become more central to economic activity, sectors, and national competitiveness, countries without substantial computing capacity face structural disadvantages in participating in this transformation. They become consumers of AI technology developed elsewhere rather than creators and innovators.
Beyond Traditional Connectivity

Lamanauskas distinguishes his concerns from previous warnings about the digital divide. Earlier connectivity initiatives focused on getting people online. That remains important, but insufficient. A nation could theoretically achieve universal broadband coverage yet still be excluded from meaningful participation in the AI economy if it lacks the infrastructure, expertise, and capital investment required to develop local AI capabilities.
This creates what economists might call a capability trap: countries that fall behind in AI infrastructure and talent development face increasing difficulty catching up, since the economic returns generated by AI investment in wealthy nations can be reinvested in further advancement. Meanwhile, developing nations must allocate scarce resources to basic connectivity while watching AI-driven productivity gains accrue elsewhere.
Systemic Implications
- Data center concentration limits where AI companies choose to invest and expand operations
- Talent migration accelerates as AI specialists pursue opportunities in developed markets
- Training data and AI models increasingly reflect the priorities and values of wealthy nations
- Regulatory power concentrates among countries that host major AI infrastructure
The opportunity divide also has cultural and political dimensions. AI systems trained predominantly on data from wealthy nations may perform poorly on tasks relevant to other populations. Development policies, healthcare protocols, and educational approaches encoded into AI systems could reinforce existing global inequalities rather than address them.
For policymakers in Africa, Latin America, Asia, and other developing regions, the challenge is finding pathways to meaningful AI participation without the massive capital requirements that currently define the industry. This might involve regional data center investment, international technology transfer agreements, open-source AI development, or other mechanisms to distribute AI capability more widely.
Without deliberate intervention, the next decade of AI deployment could resemble previous technological transitions: early advantages for wealthy nations compound over time, creating winners and losers that remain largely fixed for generations.



