$3.5-trillion hides in digital divide

Closing the mobile internet usage gap could add about $3.5-trillion to global GDP between 2023 and 2030, with more than 90% of the gains expected in low- and middle-income countries.

This projection is cited in a new GSMA Intelligence report, which examines why billions of people remain offline even where mobile broadband is available. The report was released during UNESCO’s Digital Learning Week.

Almost 96% of the global population is covered by mobile broadband, yet 3.1-billion people within that footprint do not use mobile internet. That equates to about 38% of the population.

The report, Bridging the Divide: Enhancing Digital Literacy in the AI Era, identifies digital skills, confidence and trust among the barriers preventing greater use. GSMA Intelligence produced the research in collaboration with Huawei.

Coverage is not enough

Extending network coverage therefore addresses only part of the digital divide. People also need the skills and confidence to use online services and recognise potential risks.

Artificial intelligence (AI) adds another dimension to that challenge.

Voice assistants and applications that work with images, audio and video can reduce reliance on typed commands, potentially making digital services easier to use for people with limited literacy or digital skills.

At the same time, AI increases the need for users to understand data privacy, online safety and information verification. The report highlights fraud and deepfakes among the risks people increasingly need to recognise.

The research cites World Bank data indicating wage premiums of up to 36% for generative AI literacy, exceeding those associated with broader digital and traditional AI skills.

Without targeted training, the authors warn that rapid AI adoption could deepen existing inequalities, including in places where network infrastructure is already available.

“Digital literacy can no longer be defined by static thresholds,” says Tim Hatt, GSMA Intelligence head of research and consultancy. “Connectivity alone is insufficient — capabilities, system design and trust must evolve together. To capture this multi-trillion-dollar opportunity, stakeholders must integrate layered AI literacy directly into daily livelihoods through trusted community channels.”

South Africa targets practical skills

In South Africa, Huawei has been providing free basic AI skills training to small, medium and micro enterprise (SMME) owners, university students and technical and vocational education and training (TVET) college students in collaboration with government and academic institutions.

Huawei says the programmes have reached 1,250 SMME owners and 3,140 university and TVET college students during 2026.

The GSMA Intelligence report argues that community-based programmes can help reach people most at risk of digital exclusion, including rural communities, older people, and women and girls.

One initiative cited in the research is Huawei’s Skills on Wheels programme, which includes solar-powered DigiTruck mobile classrooms. The vehicles provide hardware, connectivity and practical digital training, including AI literacy, in areas where conventional facilities may be difficult to access.

Skills on Wheels projects have trained more than 130,000 people across 21 countries since 2019.

A Kenyan government report cited in the research found that 79% of surveyed DigiTruck trainees had subsequently shared their newly acquired digital skills with family members and peers.

Gavin Allen, Huawei executive editor-in-chief, says: “Mobile classrooms have demonstrated immense value in helping underserved populations break through physical and psychological barriers to technology. We will continue working alongside local governments and industry partners to drive digital literacy, ensuring that no one is left behind as we transition into the AI economy.”

Turning coverage into participation

The report recommends treating practical AI awareness, online safety and information verification as baseline skills.

It also calls for greater use of multilingual voice and conversational interfaces in public services, expansion of mobile learning programmes, and stronger partnerships between network operators, technology companies and community organisations.

The researchers recommend measuring whether trainees can independently complete digital tasks and recognise fraud rather than focusing primarily on the number of people completing training.

The $3.5-trillion opportunity therefore depends on more than putting networks within reach. Turning coverage into economic participation requires people to have the skills, confidence and trust to use them.