Zimbabwe’s digital transformation agenda is underpinned by a policy framework that recognises the importance of inclusion.
The National ICT Policy (2022–2027) promotes equal opportunities for all persons, including women, youth, and persons with disabilities. This commitment is reinforced by the National Artificial Intelligence Strategy 2026–2030, which identifies inclusivity, non-discrimination, availability, and accessibility among its guiding principles.
The ethical AI framework is centred on human rights, safeguarding privacy, non-discrimination, due process, and dignity.
Inclusive AI development requires participatory design processes that engage affected communities, mandatory accessibility standards for persons with disabilities, gender equality measures for balanced development teams, gender impact assessments, and special provisions for youth and the elderly.
To achieve this, the Strategy adopts a phased progression from foundational capacity building to large-scale deployment and regional leadership over five years. This assumes a linear progression in which each stage builds on the previous one.
However, the International Telecommunication Union (ITU) in its Universal and Meaningful Connectivity notes that although connectivity has expanded, meaningful use remains uneven due to affordability constraints, limited digital skills, and disparities in access to quality internet services.
The Strategy is built on six interconnected pillars.
The first: talent and capacity development focuses on building AI skills through curriculum reform, specialised programmes, and centres of excellence, while recognising the diaspora’s role in knowledge transfer.
The second: infrastructure and computational capacity, addresses high-performance computing, cloud infrastructure, and data systems, and is linked to the strategy’s emphasis on data sovereignty.
The fifth pillar: research, development, and innovation aims to strengthen domestic capacity through collaboration among universities, research institutions, and industry.
To operationalise these pillars, the strategy introduces flagship initiatives. The AI Grand Challenge stimulates innovation by encouraging AI solutions to national challenges.
Project Pangolin provides a centralised data and computational platform. The Nzwisiso.ai literacy campaign promotes public awareness and digital participation. The National AI Innovation Fund supports start-ups and research.
However, when assessed against existing patterns of digital access, these initiatives risk reinforcing rural–urban and gender divides. Evidence from the ITU shows that meaningful access remains uneven across geographic, socio-economic, and gender lines.
Access to high-speed internet, data affordability, device availability, and digital literacy vary significantly between urban and rural populations. These disparities determine who can participate and who remains excluded.
Project Pangolin, which relies on centralised infrastructure, is likely to be more accessible to urban institutions. The AI Grand Challenge favours participants with existing technical expertise, institutional support, and access to digital tools.
The Innovation Fund appears to allocate resources to those who can demonstrate technical feasibility and scalability, with these capabilities concentrated in urban areas.
The strategy lacks countervailing measures, such as targeted allocations to underserved regions or mechanisms to support low-capacity entrants. As a result, the fund is likely to act as a multiplier of existing advantages rather than a tool for redistribution.
The Nzwisiso.ai literacy campaign is constrained by the same structural conditions. ITU data shows that only 35 per cent of primary schools and 49 per cent of secondary schools have internet access.
AI literacy presupposes baseline digital literacy, and where internet access is limited, AI literacy efforts risk becoming abstract and inaccessible. The campaign is likely to be most effective among already-connected populations, thus excluding rural and under-resourced communities.
Gendered patterns compound this: Women are often less likely to have access to devices, connectivity, and advanced training. The strategy does not clearly set out targeted interventions to address these disparities.
The flagship initiatives operate within existing structural conditions rather than transforming them. Those who are already connected, skilled, and institutionally supported are best placed to benefit, thereby reinforcing a cycle in which digital capability and opportunity become concentrated.
Without deliberate design to redistribute access and opportunity, these initiatives risk entrenching rather than reducing inequality.
The strategy’s phased implementation assumes foundational conditions can be established rapidly. However, evidence from the telecommunications sector reveals structural limitations.
According to data from the Postal and Telecommunications Regulatory Authority of Zimbabwe (POTRAZ), mobile and internet penetration is high, but quality and distribution remain uneven, with 4G and 5G concentrated in urban areas.
International bandwidth capacity has grown only marginally, limiting integration with global data ecosystems. The telecommunications sector is under financial pressure: operating costs have risen faster than revenues, and capital expenditure has fallen.
These constraints are compounded by unreliable electricity supply and ageing infrastructure. AI systems require stable, high-capacity infrastructure. Without it, deployment is likely to remain limited and confined to well-resourced environments.
The Strategy also emphasises data as a foundational resource, particularly through centralised platforms such as Project Pangolin. However, effectiveness depends on data quality and institutional capacity.
Evidence from the National Development Strategy 1 (NDS1) highlights constraints within Zimbabwe’s National Statistical System, particularly at ZimStats, which faced shortages of human resources, limited technical capacity, inadequate infrastructure, and constrained financial resources.
These challenges limited its ability to utilise advanced technologies and affected the quality of statistical outputs. AI systems depend on high-quality, well-structured data.
Weaknesses in data generation, management, and analysis limit AI effectiveness. Centralised platforms do not automatically resolve these underlying issues.
Computational sovereignty requires significant investment in infrastructure, technical expertise, and innovation capacity. The Strategy does not fully explain how this balance will be managed.
The Strategy correctly identifies human capital as foundational and proposes interventions to build technical skills, with NDS2 commitments to roll out training programmes in AI, cybersecurity, and data analytics reinforcing these efforts.
However, effectiveness is constrained by existing capacity gaps. The UNESCO readiness assessment highlights persistent shortages of technical expertise in machine learning and data science, as well as limited institutional capacity to deliver high-quality training at scale.
These challenges are compounded by the outflow of skilled professionals.
Digital skills hubs assume participants have reliable connectivity, devices, and baseline digital literacy, conditions that are not uniformly met, particularly in rural and low-income contexts. Training programmes may thus disproportionately benefit already-connected populations.
Expanding training capacity does not guarantee that skilled individuals will remain in the national economy or that they will find opportunities to apply their skills locally.
The ITU framework reframes readiness as meaningful participation, defined by affordability, reliability, usability, and capability. Assessed against this standard, Zimbabwe’s digital environment appears uneven and still in transition.
This gap between ambition and reality is not unique to the AI strategy. Ambitious frameworks such as Vision 2030 and NDS1 have encountered implementation constraints, with several milestones delayed or only partially achieved.
In the case of the AI Strategy, the assumption that foundational conditions can be established rapidly is not supported by the available evidence. Persistent gaps in connectivity quality, affordability, device access, digital skills, institutional coordination, and data governance remain.
The strategy seeks to scale advanced AI capabilities in an environment that has not yet achieved the digital maturity required to sustain them, risking uneven development and concentrating benefits among already-connected actors.
The question that needs to be addressed is whether Zimbabwe is pursuing AI on sufficiently developed foundations.
AI is not a starting point but a layer that relies on reliable infrastructure, inclusive access, robust data systems, institutional capacity, and public trust. Where these conditions are uneven, AI amplifies rather than resolves imbalances.
Critical is the urgent need to align it with structural readiness, shifting from rapid scaling to sequenced capability building, with infrastructure, skills, governance, trust, and rights protections treated as prerequisites.
Without alignment between ambition and operational reality, implementation will remain uneven and outcomes will fall short of intent.
This piece is part of a series on Zimbabwe’s digital sector reforms and their implications for human rights, curated by Helen Sithole. The series examines how the Cyber and Data Protection Act, the amended Broadcasting Services Act, the Postal and Telecommunications Amendment Bill, and the National AI Strategy shape freedom of expression, privacy, and access to information in Zimbabwe’s evolving digital governance landscape. Writes — Helen Sithole
