As African governments increasingly turn to automation for public services, serious questions arise about accountability and fairness. Issues such as wrongful credit removal for pensioners and misidentifications by police facial-recognition technology highlight the risks of relying on algorithms without oversight.
Dr TK Pooe, a public policy expert at the African School of Governance, emphasises that while technology is sold as a solution to bureaucratic inefficiencies, it can also obscure governmental responsibilities. “You cannot use code to hide from your constitutional obligations,” he said during a recent podcast discussion.
Research from the Tayarisha African Centre of Excellence in Digital Governance reveals that automating without adequate institutional support perpetuates existing inequalities. In his paper on algorithmic governance in Kenya, Kenneth Riany shows how algorithms mismanage public resources, often disadvantaging informal enterprises while overlooking larger corporate tax evasion. He argues that algorithmic decisions often stem from biased data, leading to unfair outcomes.
The findings suggest that relying on automated systems can shift power from citizens to opaque technologies. When public services deny citizens their rights, such as trading permits or medical supplies, the system becomes insulated from accountability. Officials can deflect responsibility by blaming the technology, while software vendors avoid liability due to proprietary protections.
With many African states purchasing expensive software from foreign firms without the technical ability to evaluate it, a lack of accountability consolidates. Ashwin Varghese’s study of Kerala, India, shows that when governments depend on external vendors, they lose control and responsiveness to citizen needs. Kerala has countered this by developing local capabilities to create public software.
The risks escalate further when automated governance intersects with security. A study by James Maisiri and his colleagues warns that state-led digital initiatives often reinforce surveillance rather than empower citizens, replicating colonial patterns of control. They argue that without proper oversight, political leaders may misuse concepts like “Ubuntu” to demand compliance while prioritising elite interests over the community’s welfare.
Proponents of public sector automation often tout efficiency and reduced corruption as benefits. However, in South Africa, Section 33 of the Constitution guarantees everyone the right to fair administrative action, and unaccountable algorithms violate this fundamental principle.
As discussions about algorithmic governance continue, the need for clear legal liability and human appeal processes becomes urgent. When technology causes harm, the state must be accountable. Outsourcing decisions to algorithms tests the resilience of constitutional protections in the digital era.
Earl Joseph is the founder of The Digital Afrikan, a media company based in Johannesburg, and a commentator on African technology policy. Stanley Moloto co-hosts The Angle Podcast, focusing on Africa’s digital innovation landscape.
