Build economies where people collaborate with AI

AFRICA

Artificial intelligence (AI) is racing fast, providing opportunities to expand capabilities to augment human intelligence, including developing scientific and expressive creativity, but not all global communities are tapping the potential of this new wave of innovations to improve lives.

For instance, there is a stark divide between low- and middle-income countries, on the one hand, and high-income economies on the other, in terms of the availability of workers with technical expertise and cognitive abilities to work with AI systems.

That is the summary of the Human Development Report 2025, A matter of choice: People and possibilities in the age of AI, released by the United Nations Development Programme (UNDP) earlier in May.

The report noted that many AI applications have not yet been widely adopted, as most countries, especially in Sub-Saharan Africa, lack the necessary elements for technology adoption, such as infrastructure, data networks and education in science, technology, engineering and mathematics.

Educational deficits

So far, the report stated, only 6% of young people in Sub-Saharan Africa meet a global standard of basic skills in mathematics and science.

According to Oxford Insights, a consulting firm that examines government AI readiness worldwide, last year’s average index rankings score for Sub-Saharan Africa was 32.7, the lowest among all the regions. Countries in the region scored between 53.94 in Mauritius and 18.58 in South Sudan.

Mauritius, South Africa (52.91) and Rwanda (51.25) were the only countries in Sub-Saharan Africa to obtain AI-readiness scores above 50. “These countries stand out as front-runners, with clear momentum in strengthening their AI ecosystems,” stated Oxford Insights.

However, despite indicators that AI technologies are increasing and even attaining stunning abilities, UNDP Administrator Achim Steiner says the primary focus should not be on what AI might achieve in the distant future, but on viewing the role of people in an AI-driven world.

According to Steiner, AI’s capacity to automate non-routine tasks has stoked fears of AI automations replacing workers.

“Our report challenges that view, as it argues that humans, the true worth of nations, are far more than the sum of the tasks they perform,” said Steiner.

AI in Africa

Even so, that appears to be the main worry for the developing world, especially in African countries, where the continent’s AI value is 2.5% of the global AI market share, according to the Global System for Mobile Communications Association, or GSMA, an industry organisation that represents the interests of mobile network operators worldwide.

Drawing insights from Eugenie Humeau, a research analyst and a manager for mobile development at GSMA, the UNDP’s report noted that many people in Sub-Saharan Africa lack the skills to use many AI tools, including affordable ones.

In a study published in 2024, Humeau and Tanvi Deshpande, a research fellow at the London School of Economics, said that, while universities offer AI-related courses, they often fail to keep pace with the industry’s needs, and students usually end up learning concepts that are no longer relevant in the AI fast race.

“In most countries in the region, the scarcity of professors with robust AI expertise and qualifications appears to be a significant challenge, as universities lack the financial resources to recruit them, effectively impacting the quality of courses offered,” stated the two researchers.

Nevertheless, in many Sub-Saharan African countries, penetration and access to AI technologies usually favour segments of the population with more financial resources. According to UNDP, whereas AI service prices have dipped globally, low-income persons in Sub-Saharan Africa can spend about 5% of their monthly income for an entry-level mobile data plan.

Access to AI

According to UNDP, access to AI tools is not merely about being online through a smartphone, as basic connectivity alone cannot unleash the full potential of the digital world or support the demands of sophisticated emerging AI technologies.

“AI applications, such as natural language processing and image recognition, depend on cloud computing facilities which require high-speed, high-quality and reliable internet connectivity,” stated the report.

Subsequently, although there is progress in data availability in some countries in Sub-Saharan Africa, on average, gaps exist in access to large volumes of data that could be used for training AI models, especially in building local language datasets.

Cultural misrepresentation

Currently, 90% of large language models are mainly trained on English data, which does not represent global regional realities.

In this regard, Humeau’s study found that the total amount of content on the internet for African languages is only 0.2% compared to 53% of English content, which is roughly 2,650 times more.

Commenting on the issue, the human development report noted that researchers on AI have observed subtle manifestations of powerplay in artificial intelligence models’ behaviour. In particular, the report stated: “Large language models trained almost exclusively on English materials can pose risks for cultural misrepresentation and even exploitation.”

However, explaining the importance of aligning AI research towards advancing human development, the report noted that inequalities exist as AI talent flows towards high-income countries, revealing widening disparities in the global distribution of AI professionals. For instance, the reports indicated that, in 2023, there were migration gains for AI professionals in high-income countries compared to losses in Sub-Saharan Africa and other developing regions.

The report predicted that a geopolitical innovation race is taking shape, but there were no indicators that Sub-Saharan Africa would matter much, as South Africa was the only country in the region cited in the race for producing patents for AI-assistive technologies.

Discrimination and exclusion

But, while there are widespread narratives about reimagining AI advancing human development in the 21st century in terms of job creation, automation and scientific progress, the report warned that developing regions, including Sub-Saharan Africa, might experience discrimination and exclusion, further entrenching existing developmental disparities.

According to the report, where many individuals lack access to affordable digital infrastructure and services could lead to exclusion from socio-economic opportunities in agriculture, health and education.

However, the report noted that some forms of AI networks were promoting child sexual abuse in various forms, even in regions with low access to digital technologies. The report noted that such abuse was rising in many countries in Sub-Saharan Africa, as most children in the region have never received advice on how to use the internet.

Digital technologies and social media networks have also been used to advance sexual harassment in Africa, with 58% of women parliamentarians being targets of sexist attacks online.

“The scope and scale of violence are constantly evolving as the rapid advance of technology provides tools that can be abused to control, silence and coerce,” stated the report.

However, although various side effects of AI technologies appear, Steiner argues that there is a need to explore their potential for development, particularly in Sub-Saharan Africa, where AI could improve education, health and agriculture and close gaps in computer use.

Shaping the workforce

Haroon Bhorat, a professor of economics and the director of the Development Policy Research Unit at the University of Cape Town, amplified such views, arguing that AI technologies are steadily shaping the South African workforce and probably elsewhere in Sub-Saharan Africa.

In a study widely quoted by the report under review, Bhorat and his associates noted that robots are not posing threats to workers. “There is a trend of people walking away from routine jobs that AI-driven robots could do, and the belief that robots are coming to take jobs has declined,” Bhorat stated.

In this scenario, UNDP recommends building economies where people collaborate with AI rather than compete against it.

According to Pedro Conceicao, the UNDP’s director of the Human Development Report Office, with the right policies and focus on people, AI can be a bridge to new knowledge, skills and ideas that can empower everyone in Sub-Saharan Africa.