Artificial intelligence (AI) has achieved the fastest diffusion rate of any technology in human history, with more than 1.2 billion users adopting AI tools in less than three years — outpacing the internet, smartphones, and electricity in speed of uptake. However, new global research by Microsoft warns that this rapid spread is far from equal, with the Global South significantly lagging in adoption and access.

According to the report, AI functions as a “general-purpose technology” — similar to the printing press, steam engine, and the internet — capable of transforming every sector of society. Yet, as with earlier technological revolutions, its diffusion depends on three key forces: frontier builders (those developing AI models), infrastructure builders (who provide computing power and connectivity), and users (who integrate AI into daily life and work). Progress, the report notes, accelerates only when all three evolve together.

While the U.S. and China dominate the AI frontier — together hosting 86% of global data centre capacity — adoption rates tell a different story. In the Global North, around 23% of working-age adults use AI tools, compared with just 13% in the Global South. The divide is widest in nations where GDP per capita falls below $20,000, a threshold that includes most least-developed countries.

Access to electricity, internet connectivity, and digital skills remain the biggest barriers to AI participation. Nearly four billion people still lack the foundational infrastructure needed to use AI effectively. In Sub-Saharan Africa, for example, 85% of those without electricity live in the region, limiting access to digital services.

The report also highlights language as a new and growing barrier. English dominates half of all online content, yet only 5% of the global population speaks it natively. This has resulted in lower AI adoption in countries where low-resource languages prevail, such as Malawi or Laos.

Despite these challenges, some nations — including Singapore, the UAE, Norway, and Ireland — have emerged as leaders in AI diffusion, proving that coordinated investment in education, digital infrastructure, and public policy can drive adoption even without frontier-level AI development.

Ultimately, the report concludes that AI’s impact will not be measured by the number of models created, but by how widely and equitably its benefits are shared across humanity.