Closing Africa’s Connectivity Gap in the Age of AI and E‑Governance

Thobekile argues that enthusiasm for AI, e-governance and digital public services is advancing faster than actual connectivity. Her point is that many people, especially rural, low-income and disabled communities, are still excluded, so the digital divide is widening rather than closing.

She introduces the session by stating that despite increased interest in artificial intelligence, e-governance and digital public services across the Global South, persistent connectivity gaps remain . She adds that while governments are rolling out AI strategies, online IDs, portals and e-services, rural communities, low-income households and persons with disabilities remain at the margins of access, often offline, stuck on 2G or blocked by high data and device costs . She further says Paradigm Initiative’s work across more than 28 countries shows a widening digital divide rather than one being bridged .

Thobekile argues that implementation at country level should involve all relevant stakeholders, not only governments and private companies. She stresses that multistakeholderism must move from global rhetoric to actual inclusive national practice.

She gives the example of Zambia, where Paradigm Initiative asked government actors how WSIS was being implemented and found that collaboration was largely between companies and government while other stakeholders were left behind . She concludes that if multistakeholderism is valued in global digital processes, it must also be applied at country level so that everyone is brought to the table .

Thobekile argues that affordable and meaningful connectivity cannot be delivered by governments alone. She says practical progress requires coordination among public authorities, private companies, civil society, technologists and communities so that international commitments translate into everyday reality.

In her closing reflections, she cites the WSIS outcome document’s emphasis on affordable broadband, blended finance, universal service funds, community networks and the role of public access facilities such as schools, post offices, libraries and hospitals . She then says connectivity issues are not just for governments but also for companies, technologists and civil society actors, and that these groups must collaborate at national level so that what is discussed in Geneva becomes a lived reality for communities .

Bridgette argues that exclusion is still severe because huge populations remain offline and governments sometimes deliberately worsen the problem through Internet shutdowns. Her point is that even where infrastructure exists, shutdowns can nullify access and deepen inequality.

She states that more than 2.6 billion people remain unconnected globally and that over half of Africa’s population is still offline . She adds that even where infrastructure exists, governments have imposed Internet shutdowns, citing Kenya and Tanzania, and notes that in Paradigm Initiative’s 2025 assessment of 29 countries, at least nine implemented shutdowns .

Bridgette argues that AI-powered identity and welfare systems can reduce rather than expand participation if they are discriminatory or inaccessible. She emphasises that these systems particularly harm elderly people, women and persons with disabilities when they are poorly designed.

She explains that many African countries are adopting AI-powered digital identity technologies, but these systems are creating challenges that reduce meaningful access, especially for persons with disabilities, women and the elderly . She gives Uganda as an example, saying research and stakeholder engagements showed that implementation of digital identity systems left elderly people out and limited their access to services . She adds that digital social protection systems have excluded elderly women in Uganda because of discriminatory and exclusionary identity systems .

Bridgette argues that affordability is a basic barrier to connectivity because expensive data makes Internet use impossible for poor households. When users must choose between connectivity and survival needs, Internet access becomes unattainable in practice.

She says data costs are very expensive in many countries and uses Zimbabwe as an example, where 1.2 megabytes costs one US dollar per hour . She contrasts this with people living on about 1.65 US dollars per day and argues that such users will prioritise other urgent needs over Internet access .

Bridgette argues that weak infrastructure remains a major reason why rural and marginalised communities stay offline. Even where policy frameworks exist, many areas still lack towers, reliable signals and stable network access.

She states that many rural communities still lack Internet access either because infrastructure is unavailable or costs are too high . She gives the example of Nigeria’s Zamfara North region, where infrastructural challenges in 2025 left many people without Internet access . Later, when discussing USF implementation, she also cites Angola as having weak signals, limited tower coverage and frequent power outages in some provinces despite laws aimed at closing the digital divide .

Bridgette argues that although many African countries have adopted Universal Service Funds, these mechanisms frequently fail in practice. The core problem is that frameworks exist on paper, but implementation is weak, delayed or absent.

She says most African countries have adopted USFs, but while some are implementing them, others are not, despite the funds being designed to cover remote areas and close the digital divide . She points to Angola, where weak connectivity persists despite relevant laws and policies . She also cites Ethiopia, where a framework was created in 2022 but the USF is still not operational , Somalia, where there is no operational USF at all , and Sudan, where repeated establishment in 2003, 2005 and 2018 still did not lead to proper functioning .

Bridgette argues that transparency is essential for USF accountability, yet many governments fail to publish even basic implementation information. Without accessible reporting, the public cannot assess whether these funds are being used properly.

She notes that in some countries governments are implementing the USF but information is not available, even though it should be publicly accessible on official websites . She gives Angola as an example where budget information and general progress reports on USF implementation are not available . She later says their biggest concern across the continent was the lack of online information and the lack of proactive disclosure by governments .

Bridgette argues that responsibility for weak USF performance does not rest only with governments. Telecom companies also undermine connectivity efforts when they fail to make required contributions to the funds.

She cites the Democratic Republic of the Congo as a case where telecommunications companies failed to make the necessary contributions to the USF . She explains that these funds are generally raised through telecom company contributions across the assessed African countries, and when companies do not contribute, rights of access to information and technology are harmed .

Bridgette argues that governments should evaluate the human rights effects of AI-linked technologies before adopting them. Her position is that such assessments can help prevent exclusion and harm before digital systems are rolled out.

She says that given the challenges outlined, countries should consider implementing human rights impact assessments before adopting or deploying technologies that integrate AI . She connects this to the broader pattern of countries adopting digital identity and AI systems without adequate safeguards or engagement .

Bridgette argues that consultation should be meaningful, inclusive and conducted well before decisions are finalised. She criticises the pattern of governments consulting only at the end of the policy process, when influence is minimal.

She says many countries adopt digital identity and AI systems without necessarily engaging or consulting people properly . She stresses that consultations must be effective and explains that too often they happen only at the end, just before implementation or adoption . She adds that consultations should include everyone and take place well in advance so that all affected groups are aware of what is being proposed .

Bridgette argues that having strategies, laws or policies is not enough if they do not produce material change. She repeatedly stresses that implementation is the decisive test of whether connectivity policies actually improve access.

In her USF discussion, she says many countries have adopted policies such as National Broadband Plans and Digital Transformation Strategies, but implementation on the ground remains very limited . In her final response, she cites Malawi’s digital economy strategy and Zimbabwe’s National Development Strategy and National Broadband Plan as promising policy frameworks, but says current reality still includes prohibitive taxes and weak access . She concludes that policy adoption is important, but it must translate into actual implementation on the ground .

Bridgette argues that while many countries struggle, some positive examples should be recognised and studied. These cases show that transparency, practical rollout and alternative approaches can improve connectivity governance.

She says South Africa ranked first among the 29 countries assessed and stood out for transparency and availability of USF information through public reporting . She adds that Namibia began Phase 1 of its USF rollout in 2025 and is doing well in expanding coverage and network connectivity . She also cites Tunisia’s use of ‘white zones’ as an alternative measure to improve connectivity in previously unserved areas, even though its USF still needs improvement .

Shumaila argues that poor connectivity should be understood as a human harm rather than merely a technical inconvenience. When digital systems fail, people can lose access to food, communication and other basic needs in both emergencies and everyday life.

She says the issue is not just about slow speed or an interrupted download, but that there is a human cost to connectivity failure . As an example, she describes flood-affected people whose biometric IDs do not work when the government announces food rations, resulting in people being denied a meal .

Shumaila argues that electricity is a foundational precondition for connectivity. Without regular power and safe charging access, particularly for women and rural communities, devices become unusable and digital participation collapses.

She describes her hometown’s electricity rationing cycle, where power is available for two hours and then unavailable for four, making it hard to charge phones . She says that when she visits, her phone becomes ‘a brick’ because she cannot charge it and may also lack data, leaving her without phone access for a week . She adds that women may have even less access to charging points, which further excludes them, and that disabled people also face additional barriers .

Shumaila argues that fiscal and financing measures can reduce barriers to connectivity. Lower taxes on devices and instalment-based purchasing schemes make handsets more affordable for people who otherwise could not buy them outright.

She says some countries reduced VAT on handsets and that this was helpful in reducing the connectivity gap, especially compared with countries that increased VAT . She also proposes device financing, explaining that even where devices become cheaper after VAT changes, some people still cannot afford them, so instalment plans can help connect women and people in rural areas .

Shumaila argues that current connectivity metrics are too crude because they count minimal or poor-quality access as full inclusion. She calls for more detailed, disaggregated measurement that reflects whether people can actually use digital services meaningfully.

She says current statistics are not disaggregated enough and may count someone with only 2G access as an Internet user, which is not useful . She recommends measuring device speed, daily use and affordability access, and insists that disaggregated connectivity data is necessary because without good information, solutions cannot be found .

Shumaila argues that USFs should not only exist but should be allocated deliberately to address inequality. She specifically suggests directing a defined share of funds towards connecting women, who face distinct access barriers.

She says some studies suggest that USF distribution should be gender-sensitive, with half of the fund used specifically to connect women and the other half for more general connectivity . She also acknowledges that USFs are struggling and that implementation countrywide is often a problem .

Shumaila argues that digital and AI-driven systems are not inherently the problem; exclusion happens when they replace older access methods entirely. She says new digital channels should be additive, while analogue or offline fallback options must remain available.

She states that there is nothing wrong with AI connectivity or AI-driven platforms in themselves . Her concern is that the ‘new AI door’ is replacing the previous door instead of being an additional one, and without fallback analogue or offline systems, entire groups are excluded . She repeats this later by saying that when a new door is introduced, it must not replace the previous one but be added to it .

Shumaila argues that AI strategies are meaningless if they do not fund the infrastructure and access conditions needed for people to use them. In her view, strategies without connectivity budgets are merely symbolic statements rather than real plans.

She says that any national AI strategy without a connectivity budget is not really a strategy but ‘just a press release’ . She adds that budgeting must include proper use of collected USF funds, which should not simply remain unused in accounts .

Shumaila argues that market incentives alone will not ensure coverage in underserved areas. She proposes making rural coverage a binding legal or contractual condition tied to spectrum access for private operators.

She notes that coverage of not only Internet but also mobile networks remains a problem even if handsets are affordable . She then proposes that private companies should accept rural coverage as part of the ‘rent’ for spectrum and says this should be a legal or contractual obligation rather than a merely ethical duty .

Shumaila argues that women cannot be treated as connected simply because a man in the household owns a device. Individual access matters because dependency gives others control over women’s information, choices and opportunities.

She responds with a concrete example: if a woman has taken a government exam and the result arrives online, a partner who controls the device could withhold the information or delete the email so that she never learns she was accepted . She adds that if a partner, boss or another person has a phone, that does not mean the woman or office is connected, and argues that targeting women for connectivity benefits everyone, whereas targeting only men does not automatically reach women .

The audience intervention challenges simplistic notions of connectivity and raises the need to be more precise about what meaningful access means for women. The underlying concern is that shared or proxy access may not provide genuine autonomy or equal participation.

An audience member questions the framing of ‘connectivity of women’ and suggests recommendations should be more specific, for example by focusing on literacy and concrete barriers rather than assuming women have distinct uncovered spaces . Although the audience member does not state the argument in exactly those words, the exchange prompts the later clarification that proxy access is not meaningful connectivity .

The audience argues that universal access debates often overlook the cost of devices, even though owning a smartphone is essential for digital communication. The point is that poor and rural populations cannot be considered meaningfully connected if they cannot afford the devices needed to get online.

One audience member says communication is a basic need and argues that discussions of universal access should focus more directly on devices . He uses the Democratic Republic of the Congo as an example, saying that in some areas people still cannot afford a smartphone even though they have a right to communicate . He further notes that some countries suffering device affordability problems are also suppliers of raw materials used to manufacture those devices .

The audience argues that discussion of USFs should not focus only on failures. Positive cases should also be identified so that others can learn from successful transparency, rollout and demand-side measures.

An audience member notes that Bridgette mentioned some success stories on USF but had not yet named them, and says these countries should be celebrated . The same intervention also asks whether any country has done well not only on network coverage but also on demand and uptake of connectivity .