ISLAMABAD, Nov 10 (ABC): Artificial intelligence (AI) offers major opportunities to improve productivity in developing economies. However, gaps in digital infrastructure, workforce skills and technological capabilities could widen economic disparities as countries adopt AI at different speeds.
The World Bank’s World Development Report 2026 estimates that 16.2% of jobs in developing economies could achieve meaningful productivity gains through AI. In comparison, the figure stands at 18.7% in high-income countries.
However, these potential gains depend on reliable internet connectivity, computing infrastructure, skilled workers and effective institutions. Countries must also develop the capacity to integrate AI into businesses and other productive activities.
The World Bank warns that without these supporting capabilities, AI could widen productivity differences between countries instead of reducing them.
Meanwhile, the report estimates that generative AI could potentially automate 4.5% of jobs in low- and middle-income countries. The corresponding figure reaches 14.2% in high-income economies.
AI productivity gains depend on digital infrastructure and skills
Prof. Dr. Muhammad Usman Ghani Khan, Director of the National Centre of Artificial Intelligence at UET Lahore, discussed these challenges with Wealth Pakistan.
He said the next global productivity divide could emerge from differences in countries’ ability to use AI effectively. Access to AI models alone would not determine which economies benefit most.
“Access to an AI tool is increasingly becoming easier, but productive adoption requires much more. Countries need reliable connectivity, computing infrastructure, skilled people, usable data and organisations capable of redesigning the way work is performed,” he said.
“If those foundations are missing, AI may remain a consumer technology rather than becoming a productivity technology,” he added.
According to Khan, developing economies must strengthen the systems that allow businesses and workers to use AI productively. Otherwise, access to advanced technology may not translate into higher economic output.
Internet connectivity gap limits AI adoption
Internet connectivity data highlight a major challenge facing developing economies.
According to the International Telecommunication Union (ITU), approximately 6.3 billion people worldwide are using the internet in 2026. However, around 2 billion people remain offline.
The gap between high-income and low-income economies remains substantial. Internet usage has reached 93% of the population in high-income countries, compared with only 32% in low-income economies.
Regional differences are equally significant. In Africa, only 42% of the population uses the internet. Meanwhile, connectivity reaches 41% across least developed countries and 43% in landlocked developing economies.
These disparities could slow the adoption of AI-assisted production, business services and decision-making. Moreover, unreliable connectivity can prevent workers and companies from using digital tools consistently.
A joint study by the International Labour Organization and the World Bank examined AI’s potential effects across 135 countries. It found that developing economies face relatively low exposure to outright AI automation but have substantial opportunities for worker augmentation.
However, the study warned that inadequate digital infrastructure could cause technological disruption before countries realise the productivity benefits.
Khan said this imbalance was particularly important for developing economies seeking stronger economic performance.
“A worker may theoretically benefit from AI assistance, but that benefit cannot materialise consistently if internet access is unreliable, computing is expensive or the workplace has not digitised its processes. Productivity gains require an ecosystem, not simply software,” he said.
AI investment concentration poses challenges for developing economies
The global investment landscape presents another challenge as capital increasingly flows towards a limited number of countries and sectors.
UN Trade and Development (UNCTAD) warned in 2026 that investment in AI and digital infrastructure was becoming increasingly concentrated. Consequently, developing economies unable to attract sufficient investment could fall further behind technologically advanced countries.
In September, UNCTAD also reported that fewer than one-third of developing countries had adopted national AI strategies.
Furthermore, computing capacity, finance, data and technical expertise remained concentrated among a relatively small group of economies and companies.
These trends underline the importance of strengthening domestic technological capabilities. Developing countries also need investment conditions that encourage businesses to adopt AI across different economic sectors.
Without broader access to infrastructure and expertise, AI adoption could remain concentrated among large companies and technologically advanced industries.
Agriculture, healthcare and manufacturing offer AI opportunities
Khan said developing economies should initially focus on practical AI applications that deliver measurable productivity improvements.
Rather than investing heavily in expensive computing infrastructure for frontier AI models, countries could adapt existing technologies to local economic needs.
“Agriculture, healthcare, education, manufacturing, public administration and financial services all offer opportunities. The objective should be to adapt AI to local languages, datasets and operational problems rather than trying immediately to compete in building the largest models,” he said.
The World Bank similarly argues that developing countries do not need to develop the most expensive frontier models to benefit from AI.
Instead, they can adapt available AI technologies to their languages, institutions and development priorities. This approach could help countries improve productivity while gradually building technical expertise and digital infrastructure.
Moreover, locally adapted AI applications could make advanced technology more useful for businesses, workers and public institutions.
Wider AI adoption essential to prevent productivity disparities
Khan said the central policy challenge was ensuring that AI capabilities reached a broad range of businesses and workers.
He warned that concentrating advanced technology in major cities, leading companies or developed economies could increase existing productivity differences.
“If AI capability becomes concentrated only in leading firms, major cities or technologically advanced countries, productivity gaps can widen very quickly. The real development objective should be to spread useful AI capacity across businesses, workers and public institutions,” he said.
For developing economies, the global AI transition therefore presents both opportunities and challenges. While AI could improve economic performance, unequal access to essential resources could deepen existing disparities.
Ultimately, the long-term benefits will depend on countries’ ability to turn AI adoption into sustained productivity growth.
Building reliable infrastructure, developing workforce skills and strengthening institutional capacity will remain essential. These measures could help developing economies benefit more widely from AI-driven technological progress.
