Ghana's Ghanacard covers approximately 85% of the adult population yet systematically excludes Fulani pastoralists through discriminatory citizenship scrutiny, exposing a foundational tension at the heart of the country's digital welfare project. This paper examines how Ghana's digital identity infrastructure, anchored in the Ghana Card and progressively integrated with biometric, administrative and social programme databases, is reshaping the design and delivery of social protection.
Drawing on institutional analysis of policy documents and on primary interviews conducted in early 2026 with officials from the National Identification Authority (NIA), the Ministry of Gender, Children and Social Protection (MoGCSP), the Data Protection Commission and Fulani community leaders, the study situates AI-enabled welfare administration within a broader political economy framework organised around state capacity, distributive justice and algorithmic governance. The paper examines whether AI-supported digital identification enhances targeting efficiency and inclusion or reproduces novel forms of exclusion through data gaps, biometric failures and opaque decision-making.
Key findings include a 14 to 15 percentage-point increase in NHIS uptake linked to the LEAP 1000 pilot, alongside documented ethnic marginalisation, weak data protection enforcement and a digital transparency deficit that privileges urban elites.
The analysis is constrained by scarce data on actual AI usage. No public dataset details algorithm accuracy or bias in Ghana's targeting. We therefore infer implications from process data in terms of how ID registration affects programme inclusion and from analogous settings. The approach is not statistical; rather, it aims to paint a holistic picture grounded in policy analysis and qualitative evidence.
The study contributes a Global South political economy perspective on digital state formation, data justice and algorithmic welfare governance and offers concrete policy recommendations for ethical, accountable and inclusive deployment of AI within emerging digital welfare systems.
Particular attention is paid to implications for marginalized populations, informal sector workers and digitally invisible citizens
Policy recommendations are offered to guide ethical, accountable and inclusive deployment of AI within emerging digital welfare systems.
