The development of artificial intelligence (AI) in today’s world is remarkable – projections show that the global revenue of the AI market is set to grow by 19.6% each year, and reach $500 billion by the end of 2023. Tech giants are investing billions into its development, and the technology is evolving exponentially with recent groundbreaking advances. Despite the enthusiasm around AI technologies, experts have repeatedly highlighted the importance of developing new AI tools responsibly, to avoid biases and ensure safe development. Despite ongoing debates regarding AI regulation, the challenges faced by the Global South are largely missing from the conversation, and there is a void in the studies discussing AI’s potential role in deepening digital divides between developed and developing economies.
The rapid advancement of AI has the potential to exacerbate the existing levels of inequality and social division between the Global South and Global North. Since levels of AI development, adoption and usage are dependent on digital infrastructure, much remains to be done before developing economies, marked by lower levels of network readiness, can harness the full benefits of AI. To exemplify the potential deepening digital divide between developed and developing economies, we present three reflections and cases.
The first example can be seen when analyzing the future of labor in the era of technology. Considering that investments in AI are mainly from high-income countries, the technologies being developed aim to address those countries’ needs in their specific context. Professor Daron Acemoglu (MIT) has pointed out that capital-intensive innovations developed in advanced economies might not be particularly useful in poor economies where labor is abundant and capital is scarce. He further states that in countries like India or Brazil, which face higher relative costs of capital, the appropriate technology is one that actually prioritizes labor-intensive production methods. However, because advanced economies have no reason to invest in such labor-intensive technologies, the trajectory of technological change will increasingly disfavor poor countries.
The economic structure of many countries in the Global South is reliant on labor intensive industries. Many of these countries also have a significant advantage as labor costs are considerably lower compared to capital investments. For instance, more than half of Africa’s population is of working age. Therefore, capital-intensive innovations might undermine this advantage and further contribute towards increasing unemployment in low income countries. Further, capital-intensive AI innovations are difficult to implement and operate in many developing countries due to limited financial resources and lower levels of digital skills. Consequently, a more suitable approach would be to diversify the development of AI technologies in order to cater to the developmental needs of the middle and low-income countries, allowing them to leverage their labor cost advantage and promote economic growth, thus contributing to a more equitable landscape.
Second, research and publications in the field of Explainable AI are also scarce in developing economies. A study conducted by Cornell University in the field of Explainable AI (XAI) in developing countries found that although XAI is a rapidly growing area of research, most of the work has focused on contexts in the Global North, and little is known about if or how XAI techniques have been designed, deployed, or tested with communities in the Global South. After a thorough examination, the authors identified only 16 papers on that topic, that targeted a wide range of application domains. From these works, only three papers engaged with or involved humans in the work, and only one attempted to deploy their XAI system with target users. The scarcity of research on the XAI field in the Global South is one additional concern to the responsible development of AI, especially in countries where digital literacy is still a challenge to be addressed by governments.
As a third example, recent debates have emerged on the use of AI for healthcare applications. Experts have pointed out that the publicly available databases for training AI algorithms are, for the most part, from developed countries, and do not reflect the reality of the developing world. Renata Prôa, a data analyst at Hospital Albert Einstein (Brazil), explains that most publicly available datasets on skin cancer originate from Europe, North America, and Oceania, regions which have a predominantly white population. Considering that machine-learning processes are based on the analysis of such datasets, if AI is developed based on Caucasian data and used in countries with a different phenotypic conformation, it could lead to further bias and discrimination and imprecise assessments.
Based on these examples, digital readiness levels in the Global South must be a priority area in order to ensure that the benefits derived from groundbreaking technological development are equally distributed. Otherwise the digital divides between developed and developing economies are at risk of continued deepening as frontier technologies continue to evolve.
According to Professor Daron Acemoglu, AI-driven divergence between the rich and developing worlds is not inevitable, but needs to be addressed. Developing countries must urgently adopt measures to ensure they are not left behind in the technological transformation. These include improving network readiness, fostering R&D initiatives by public and private players, adopting policies to develop publicly available databases (to allow AI development suited to the local context on public healthcare and education, for instance), and implementing legal and regulatory frameworks to mitigate the risks and biases of AI. Similar discussions have also been surfaced by the 2021 edition of the Network Readiness Index, which focused on the role of technology as an equalizer in the period of post-COVID recovery.
The future of our digital society will largely be shaped by the development and advancement of emerging technologies. Global benchmarking tools such as the Network Readiness Index can help determine the proper courses of action to ensure that the economic and social benefits of this enhanced digital revolution are distributed equally across the globe.
The 2023 edition of the Network Readiness Index, dedicated to the theme of trust in technology and the network society, will launch on November 20th with a hybrid event at Saïd Business School, University of Oxford. Register and learn more using this link.
For more information about the Network Readiness Index, visit https://networkreadinessindex.org/
Matheus de Souza Depieri is a Brazilian researcher and lawyer, currently pursuing an LL.M. (Master of Law) at the University of Cambridge – King’s College. He is also Associate Editor of the International Review of Constitutional Reform (IRCR), Director of the Center for Comparative Constitutional Studies (UnB), and researcher at the Research Group on Civil Procedure, Access to Justice and Protection of Rights (CNPq/UnB). During his Fellowship at the Portulans Institute, Matheus’ research focuses on the impacts of the usage of Artificial Intelligence in the Judiciary and the future of technology in the legal field.
Samantha Msipa is a passionate researcher and has actively engaged on various projects on law and technology throughout her career. She has a Bachelor of Laws degree as well as a Master of Laws in International Commercial Law from the University of Johannesburg. Her affiliation with Research ICT Africa as an AI Research Fellow allows her to delve deeper into the complexities of AI and its implications on the African continent.