The advancement of semiconductors, supported by the availability of data and lower computing cost, has led to a significant increase in the commercial adoption of AI. If I told you that the next emerging AI markets had millions of young, tech-savvy consumers and little to no old legacy systems to strangle them, you may think they’re in the US or Europe. And you’d be wrong.
Think beyond the Organization for Economic Development (OECD) and beyond China and India, whose market size is projected to exceed $61B and $17B by 2027. Think instead of Nepal, Nigeria, Bangladesh, Ghana and many more. These are the Emerging AI Markets (EAIMs). UN projections for 2030 place their total population at nearly 4.2 billion, far larger than the entire OECD. With these economies building infrastructure and expanding their healthcare systems, they can build in AI from day one. Sooner or later, it will deliver rapid productivity growth.
The signs are already there: Nepal-born AI enterprise software company Fusemachines is preparing to be listed on the NASDAQ. This reminds me of the 1990s when Infosys and Wipro put India on the global technology map. From Ghana, mPharma has crossed borders by acquiring Kenya’s Haltons pharmacy chain, while in Bangladesh, bKash got a $250 million investment from SoftBank Vision Fund II. This clearly shows how EAIMs are marking their space in the global AI-driven economy.
The stakes are high: Artificial intelligence has the potential to reverse global inequality — or to make it worse. This article explores the opportunities for investors and tech firms who seek the former.
What Are Emerging AI Markets (EAIMs)?
EAIMs are countries like Nepal, Nigeria, Bangladesh and Ghana, characterized by young, tech-savvy populations and minimal legacy systems, poised to lead the global AI-driven economy through infrastructure development, talent cultivation and innovative solutions.
Beyond Outsourcing
Many Silicon Valley companies may see EAIMs as a place for offshoring back-office operations. But that’s just scratching the surface. These economies are competing head-to-head against advanced markets and are building solutions in their own backyards that can help solve global problems.
For example, the AI company Fusemachines’ listing on NASDAQ suggests Nepal could become a hub for enterprise-level software expertise. In the same manner, companies like CloudFactory, which is a Dolma portfolio company alongside Fusemachines, provides data solutions globally from Nepal and Kenya. Emerging economies are no longer just consumers of digital technologies. Instead, they create new markets and influence how innovation evolves.
EAIM companies are also helping to transform domestic industries. In Ghana, mPharma manages more than 243 pharmacies and clinics and serves 40,000 patients a month, which has helped to cut drug prices by up to 30 percent. In Nigeria, where only about 3 percent of adults have health insurance, aYo Holdings’ mobile-first microinsurance platform has issued more than 20 million policies. Millions of farmers in South Asia use tools such as Plantix to diagnose crop diseases, while Qure.ai has expanded AI-powered radiology to more than 50 countries experiencing a shortage of medical professionals.
These companies show the commercial viability of EAIMs. But now, here’s our next challenge: building enough talent to sustain the momentum.
Solving the Education Gap Without Lowering the Bar
A common question about these markets is whether EAIMs can produce enough skilled AI talent. Based on my personal experiences, they can, and the model is already operating at scale.
Fusemachines built its business model around education. Through its AI Fellowship and Foundation courses, engineers in Nepal and other emerging economies complete a six-month, project-based program covering core frameworks such as TensorFlow and PyTorch before they join the company. This way, a steady stream of professionals are both technically skilled and ready to work. More than 800 AI engineers have received training thus far.
At the same time, CloudFactory looks at things from a different angle. Its distributed workforce runs large-scale data annotation and human-in-the-loop pipelines that are certified to international standards like ISO 27001. This reassures global clients that teams based in EAIMs can deliver critical AI infrastructure tasks to the same standards of quality as teams in Silicon Valley.
Similarly, in Nigeria, Decagon provides a rigorous coding fellowship that combines in-person job placement with advanced training. The results are remarkable: graduates from cohorts with an acceptance rate of less than 1 percent experienced a 100 percent job placement rate, a 100 percent loan repayment rate and a 410 percent salary bump. Decagon demonstrates how closely integrating education-to-employment models can transform careers and national talent pipelines.
In short, EAIMs are resolving the long-standing global AI talent-pipeline shortage. Without sacrificing or compromising on talent, they’re scaling rapidly, training people on industry-standard tools and deploying into global projects.
Infrastructure: Where the Compute Will Live
Talent and services are only half the story. AI data needs somewhere to live. Where will those data centers be built, and how can we power them cost-efficiently and sustainably?
China shows a perfect example in the Tibetan Plateau. The Yajiang-1 site, developed under the country’s “Eastern Data, Western Computing” push, sits at 3,600 meters and uses altitude, along with a cold climate and rivers, to reduce cooling loads.
Globally, the demand is staggering. According to McKinsey, AI-ready infrastructure is expected to grow by more than 30 percent annually, with capital investment likely to exceed $5 trillion by 2030. And this build-out won’t be limited to the US or Europe just because governments are demanding that data be stored locally.
All the components needed to satisfy this demand are found on the Himalayan plateau. Nepal and Bhutan have vast hydropower reserves, solar potential, cool mountain climates, glacial rivers and low electricity prices. With affordable, clean energy and natural cooling, these nations have the potential to house some of the most economical and environmentally friendly data centers in the world.
Other EAIMs are leading the way. In Kenya, a Microsoft-G42 initiative is building a geothermal-powered data center, starting at about 100 MW —enough electricity to power a mid-sized city— with expansion potential toward 1 GW, which is ten times larger and on par with the output of a full-scale nuclear or coal power plant. In Southeast Asia, electricity consumption from data centers is expected to rise from 9 TWh in 2024 to 68 TWh by 2030, driven largely by cooling in tropical climates. The trend is clear: We need new locations and smarter energy choices.
The bottom line is that AI infrastructure doesn’t have to be limited to northern Virginia or Frankfurt, Germany. A foundation of global computing could be situated in EAIMs with a wealth of renewable energy sources and favorable climates. If these projects are implemented to international standards and stakeholders ensure that local populations are treated as inclusive beneficiaries, the socioeconomic transformation will be huge.
The Risk of Inequality and How to Avoid It
Every industrial revolution has had its own winners and losers. Though developed economies have rushed ahead, the Global South has often been excluded from the process of wealth generation. The first and second Industrial Revolutions led to the concentration of wealth and technology in Europe and North America, while colonies and the Global South mostly supplied raw materials and cheap labor pools.
The third, digital revolution further widened the divide, and today around 2.6 billion people remain offline. We need to ensure that the fourth industrial revolution (i.e., the AI revolution) doesn’t widen the gap further. Unless we act differently this time, AI risks repeating the same pattern and could reinforce inequality which will ultimately lock billions out of the next wave of prosperity.
This time, however, could be different. Poorer nations can compete on a level playing field without an infrastructure deficit (e.g. roads, rail, ports) hindering their inclusion. All an EAIM company needs to compete globally today is brains and broadband. Brains are distributed equally, and broadband is catching up in developing markets.
But if EAIM entrepreneurs, talent and data infrastructure are excluded from investment pipelines due to historical bias, the economic value of AI will be concentrated in established hubs. Capital, computing and jobs will stay in developed markets, while EAIM economies are reduced to algorithm takers rather than algorithm makers. That outcome would widen the gap between rich and poor nations.
The irony is that many of the most impactful solutions are already being born in EAIMs, where service cost reductions from technology are more critical. Such innovations include affordable healthcare, crop diagnostics and insurance access. These innovations will, in turn, benefit developed economies.
At the macro level, the opportunity is even larger. IFC estimates that the digital economy could create up to 230 million new jobs in Africa by 2030. McKinsey projects multi-trillion-dollar investment in AI infrastructure globally on the same timeline. The scale of what is at stake is hard to overstate.
If we invest with both inclusion and commerce in mind, the story changes. AI can create middle-class digital jobs at scale, solve local challenges in healthcare and agriculture and enable EAIMs to contribute to, and not be passengers in, the AI revolution.
This is the choice before us. Either AI deepens global inequality, or it becomes the first industrial revolution that spreads its benefits more evenly.
A New Playbook
EAIMs are no longer a footnote in the AI revolution; they’re an essential part of it. The challenge now is for investors, corporations and governments to seize the opportunity.
The entry points are clear. Businesses should partner with companies such as those featured here that have already proven EAIMs can scale globally. Pilot deployments in healthcare, fintech and agriculture, where financial and social benefits coincide. Back the infrastructure that will make this possible: data centers, regional AI hubs and the sustainable energy that powers them. Above all, as investors, we must broaden the horizon of our pipelines.
At Dolma Fund Management, we have been investing in this space for more than a decade. From what I’ve seen on the ground, the infrastructure and talent of EAIMs can produce long-lasting effects and competitive returns. Investing in EAIMs is commercially smart and historically necessary. AI will either entrench global inequality, or it will build the foundation of a fairer future. The choice is ours, and the time is now.
