Why 5G, AI and IoT Alone Will Not Bridge the Digital Divide

The digital divide is no longer just about who has internet access. In developing economies, the sharper divide is between people who can turn connectivity into better healthcare, education, finance and livelihoods, and those left with weak networks, expensive data, limited skills and little institutional support. As 5G, AI and the Internet of Things spread, the risk is that smarter infrastructure could widen old inequalities instead of erasing them.

In "Bridging the Digital Divide in Developing Economies Through Intelligent Connectivity (5G, AI and IoT)—Insights from a Structured Literature Review," published in Telecom, Laurence Banda and Etienne Alain Feukeu of the University of South Africa examine whether this next generation of connectivity can deliver genuinely inclusive development. Their review of 63 peer-reviewed studies finds that the technology has real potential, but its impact depends on far less glamorous foundations: affordability, digital skills, governance, institutional capacity and investment.

The Digital Divide Is No Longer Just About Getting a Signal

Digital exclusion in developing economies is increasingly defined by quality, affordability, capability and use rather than a simple connected-versus-unconnected divide. The study distinguishes between access, skills and usage gaps, noting that weak infrastructure can coexist with high prices, low digital literacy and limited participation in productive digital activities.

Sub-Saharan Africa illustrates the difference between coverage and meaningful adoption. The review cites evidence that around 60% of the population lives within mobile broadband coverage while only 28% uses mobile internet services. Women in the region are also reported as 37% less likely than men to use mobile internet, showing how infrastructure, affordability and social barriers can compound one another.

Cost remains a particularly powerful barrier. In Sub-Saharan Africa, the study reports that 1GB of mobile data can consume 5–15% of monthly income for low-income households, compared with 1–2% in high-income economies. Even where networks exist, the price of devices and data can therefore convert nominal coverage into a persistent usage gap.

The value of connectivity increasingly lies in what people can do with it. Messaging and voice access are not equivalent to participation in digital finance, online education, telemedicine or e-commerce. A country can improve headline connectivity indicators while leaving the deeper economic divide largely intact.

5G, AI and IoT Only Matter When They Work Together

The study treats 5G, AI and IoT as interdependent rather than separate technologies. In its framework, 5G enables reliable high-speed communication, IoT captures data from physical environments, and AI interprets those data for automation and decision-making. The development effect comes from their interaction, not from deployment of any one technology in isolation.

The convergence creates tangible opportunities. In agriculture, sensor networks can collect field data while AI supports resource optimization; in healthcare, connected systems can enable remote monitoring and AI-assisted diagnosis; and in education, intelligent platforms can support adaptive and multilingual learning. The review also highlights applications in financial services and disaster management. However, each application carries its own inclusion problem. The study cites Hello Tractor as an example of AI- and IoT-supported agricultural inclusion, but notes that smartphone access, data connectivity, rental costs and maintenance can exclude poorer farmers. Similarly, AI-enabled financial services may expand credit access while creating risks of algorithmic discrimination and weak data protection.

Technological sophistication does not eliminate underlying scarcity. A smart agricultural platform still depends on affordable devices and connectivity; telemedicine still requires networks, electricity and institutional trust; and AI-supported services still need data governance. Intelligent connectivity can reduce barriers, but it can also reproduce them in more technologically complex forms.

There Is No Single Digital-Inclusion Playbook for the Global South

The paper compares Sub-Saharan Africa, Southeast Asia and Latin America and finds very different development constraints. In Sub-Saharan Africa, the dominant barriers are infrastructure deficits, affordability and weak institutional capacity. Southeast Asia combines strong urban adoption with rural exclusion, while Latin America has comparatively higher connectivity but persistent territorial and socioeconomic inequality.

The regional contrast is strategically important. Estimated internet penetration cited in the study was 34% in Sub-Saharan Africa, 72% in Southeast Asia and 75% in Latin America for 2023, while mobile broadband coverage reached 60%, 85% and 92% respectively. The numbers underline why policies designed for a relatively connected Latin American city cannot simply be transferred to an African rural region with weak electricity and limited fibre backhaul.

Sub-Saharan Africa faces some of the most severe infrastructure constraints. The review notes that 5G base stations can consume three to four times more energy than 4G while roughly 600 million people in the region lack reliable electricity access. It also points to limited fibre networks, high device costs and regulatory bottlenecks, warning that near-term 5G benefits may concentrate in urban and peri-urban areas.

Southeast Asia presents a different challenge: relatively strong mobile adoption can coexist with fragmented regulation, difficult geography and urban-rural skill gaps. Latin America, meanwhile, shows how advanced digital services can remain concentrated in affluent areas even when overall connectivity is high. The study thus argues for context-specific digital strategies rather than one-size-fits-all technology rollouts.

Governance Will Decide Whether Intelligent Connectivity Includes or Excludes

The study proposes integrated digital governance rather than fragmented policies for individual technologies, including adaptive spectrum management, rural connectivity incentives, regulatory sandboxes, public-private-community partnerships, digital equity strategies and stronger data governance.

For businesses, the opportunity lies in designing models that work under real constraints rather than assuming mature infrastructure. Shared networks, network-as-a-service models, local AI and IoT applications, and cross-sector collaboration between telecom operators, energy providers and public institutions could extend services into underserved markets while lowering deployment barriers.

However, the risks are substantial. The authors warn of algorithmic bias, surveillance, dependence on foreign cloud and technology providers, infrastructure lock-in, widening gaps between users with advanced connectivity and those with basic or no access, and the environmental costs of energy use and electronic waste. Intelligent connectivity could therefore produce a more stratified digital economy if governments prioritize adoption over inclusion.

The evidence base also requires caution. Only 10 of the 63 peer-reviewed studies reviewed, 16%, focused specifically on developing economies, while 84% came from developed-economy contexts. The authors also acknowledge English-language and database bias, substantial differences within regions, early-stage deployment of many technologies, largely descriptive case evidence and the absence of primary data collection. The limitation is itself revealing. Policymakers are being asked to make expensive decisions about 5G, AI and IoT in regions where the empirical evidence remains comparatively thin.

The study calls for country-level validation, stronger impact measurement, research on algorithmic bias and digital sovereignty, low-resource business models, local-language AI, energy efficiency and open standards.

Closing the digital divide will require more than deploying advanced networks. Intelligent connectivity can improve agriculture, healthcare, education, finance and public services, but its impact will depend on whether people can afford it, understand it, trust it and use it productively.