Theyre the A.I. exceptions making an impact in the Global South today that must prove to be the rule tomorrow. In Kenya, Safaricom and Commercial Bank of Africa have been using A.I. to review online loan applications from remote areas; to predict the probability of defaults; and to provide small loans to over 20 million Kenyans. In Mexico, Clinicas de Azucar has deployed the technology to gather a variety of patient data; analyze that data; and make recommendations to over 150,000 Mexicans who run the risk of becoming diabetics. In Brazil, Revelo is using A.I. to study online education and job platform data to deliver upskilling recommendations that will improve many Brazilians employability.
Despite these promising beginnings, the use of A.I. in the countries of the Global South is still uncommon (other than in China, one of the worlds A.I. co-leaders). The incumbents use A.I. infrequently; too few digital ventures work on developing applications; and the technology is a long way from attaining scale. Just 7% of the companies in the Global South use A.I. at present, compared to nearly 60% in China, according to recent data.
If the Global Souths companies continue to trail those of the developed world in adopting A.I., they run the risk of becoming marginalized even at home. In many sectors, foreign rivals will sooner or later overtake the domestic market leaders by using A.I. To remain competitive, the incumbents must play the role of A.I. national champions, and become the prime movers in the use of the technology in their home markets. Not only will doing so enable the incumbents to unlock additional value as first movers but also, it will help them retain their market leadership.
Importantly, becoming A.I. champions will allow the Global Souths leaders to catalyze the creation of domestic markets for the technology. Developing local A.I. markets is critical if innovation, technology development, and application-creation are to flourish there. In countries where A.I. is increasingly being adopted, we find, cutting-edge applications are being developed quickly. These markets develop a healthy appetite for the latest applications, making it viable for entrepreneurs to invest in developing them.
Chinas A.I. firms, for example, have benefitted from a large domestic market. Nearly 60% of IT professionals in China say their organizations use A.I., compared with the U.K.s 26%, the U.S.s 25%, Australias 24%, and South Koreas 22%, according to IBMs Global A.I. Adoption Index. And the number of Chinese A.I. firms rose to around 1,189 last yearsecond only to the U.S.s 2,000-plus firms.
One of those Chinese firms is the Hefei-based iFlyTek, which has invested heavily in developing A.I.-enabled speech technologies. As the demand for speech-based applications has increased, iFlyTeks products have made inroads in Chinas communications, healthcare, education, and services sectors. The impetus provided by a growing domestic market has enabled the $2.875-billion venture to continue investing in technology development; its R&D expenditure grew at 28% between 2014 and 2021, to reach $150 million last year.
Driving A.I.s use is critical to tackle the three challenges that hinder its adoption in the Global South: the lack of talent, infrastructure, and local A.I. firms. For one thing, most emerging markets suffer from a dearth of A.I. talent. According to ITU data, the percentage of people with computer skills was 19% in the Global South, compared to 51% in the Global North. The low percentages arent offset by the developing worlds large populations; according to the Tortoise Global AI Index, programmers in the Global South accounted for less than 5% of the changes made to A.I.-related files (commits) on GitHub, a proxy for the quality of the talent working in the field.
For another, and relatedly, the Global South suffers from a lack of infrastructure to support A.I. development. According to the Tortoise Global AI Index, emerging markets collectively hosted fewer than 20 of the worlds 500 fastest (non-distributed) supercomputers, which are vital for developing and testing A.I. algorithms, in 2021.
Finally, low income nations havent spawned many A.I. firms, so the market leaders that wish to adopt the technology arent able to do so easily. Only India ranks among the worlds top 10 A.I. adopters, although its $3 billion market was less than a sixth the size of Chinas $20 billion market in 2021.
The Global Souths national champions, our studies show, can take the lead in overcoming these hurdles by deploying a three-step approach.
Step one: Start
The A.I. national champions must begin by identifying business processes and organizational activities in which they can gain a competitive edge by using A.I. Going after the low hanging fruit first will fuel success and enlarge appetites.
In India, for instance, Craftsman Automation, which started as a small-scale machining operator in the late 1980s, has become a leading supplier of automotive parts by deploying A.I. A decade ago, it started using the technology to develop self-optimizing machines and automate quality control, working with American giants such as Rackspace Technology and IBM as well as European ones like Germanys Carl Stahl.
If the domestic A.I. ecosystem is still nascent, the national champions have no choice but to work with global companies. Doing so will be expensive initially and the gestation period may be long, but they must take the leap immediately to gain an advantage. Only by using A.I.-based solutions can the national champions lay the foundations of a local ecosystem that catalyzes the growth of local A.I. start-ups and creates opportunities for homegrown talent.
By pioneering A.I.s use, Craftsman Automation has catalyzed demand for the technology in Indias automobile industry. That has led to the birth of over 100 A.I. startups that focus on developing manufacturing applications. Indias talent pool has also benefited, with business sponsoring post-graduate A.I. programs at several of the countrys leading universities.
Step Two: Structure
The Global Souths A.I. national champions must work with governments as well as business to create effective structures for A.I. ecosystems and develop national infrastructures. They must establish standards that will help companies build on each others efforts, so they dont have to start every digital initiative from scratch.
The digital public goods created by properly structured ecosystems will hasten the next phase of A.I.s adoption by attracting more users and ventures. Indias National Payments Corporation of India, for instance, was created by the countrys central bank and a consortium of local banks; it launched its Unified Payments Interface (UPI) in 2016. A real time system that facilitates person-to-person and person-to-merchant transactions, it runs as an open-source application programming interface atop the countrys inter-bank electronic funds system. By September 2022, the number of Indian banks using the UPI had risen to 358, transaction volume had reached 6.7 billion a month, and their value exceeded U.S. $140 billion. In fact, India accounts for nearly 40% of all the digital payment transactions in the world today.
Moreover, many fintech start-ups have sprung up in India, and theyre using the UPI to increase their reach. Based on their growing understanding of the customer base, these firms are developing A.I.-based applications that provide personalized and value-added financial services, thereby accelerating the sectors growth.
Step Three: Switch
Whenever the Global Souths national champions help catalyze A.I. ecosystems, it will spur the creation of numerous benefits at home. As a domestic A.I. ecosystem develops, it will trigger a virtuous cycle of greater talent creation, lower development costs, and increased acceptance of the technology, which will create more opportunities and boost the national champions competitiveness.
The incumbents will also be able to stop depending on foreign technology-providers and switch to indigenous ecosystems. Many will find it easier to work with local A.I. firms, which tend to understand customer needs and local markets just as well as the incumbents do. Doing so will also lower costs because of the use of local talent and that, in turn, will enable the domestic A.I. market to grow faster.
For example, the South African macadamia processing co-operative, Mayo Macs, decided to use A.I. to detect disease in macadamia trees, so its member-farmers could increase their output. It has developed solutions by teaming up with Aerobotics, a Cape Town-based start-up that specializes in providing crop health data. Its drones take photographs of fields from the air; the start-up uses A.I. to process the aerial images; and it identifies tree infections before they become visible to human eyes. Such local solutions for local problems allow the Global Souths economies to double their benefits from the technology, by turning into both A.I. developers and A.I. users.
Focusing on A.I. will allow the national champions to become more competitive both at home and abroad, especially in the fast-growing markets of the Global South. Given the similarity of emerging markets economic, social, and cultural contexts, these companies are better suited than Western multinationals to use A.I. to develop novel business models that leapfrog traditional ones and reach underserved markets. Doing so will not only drive business towards an A.I.-focused future, but it will also help boost prosperity and shrink poverty in the Global South by unleashing A.I.s full power.
Read other Fortune columns by François Candelon.
François Candelon is a managing director and senior partner at BCG and global director of the BCG Henderson Institute.
Maxime Courtaux is a project leader at BCG and ambassador at the BCG Henderson Institute.
Gaurav Jha is a consultant at BCG and ambassador at the BCG Henderson Institute.
Some companies featured in this column are past or current clients of BCG.
