Poor data quality and weak governance frameworks could undermine Africa’s growing artificial intelligence ambitions unless governments and institutions prioritise building stronger data foundations before deploying the technology, a technology expert has warned.
The caution comes as several African countries, including Nigeria, Kenya, and South Africa, roll out national AI strategies and accelerate investments in artificial intelligence to drive innovation, improve public services, and strengthen economic competitiveness.
Nigeria officially launched its National Artificial Intelligence Strategy in April 2025 in Lagos. Since then, the government has moved into implementation with several key developments.
According to the Founder of UK- and Nigeria-based data consultancy Metaheuristic Limited, Toye Apampa, many organisations are treating AI as a technology challenge when the real obstacle lies in the quality and management of the data that powers these systems.
“By the time a leadership team is asking how to deploy AI, the data underneath is usually in worse shape than they realise,” Apampa wrote in an article on Africa’s AI readiness made available to The . “The clean dashboards in the boardroom do not survive contact with what is in the database.”
Governments and institutions across the continent have shown growing interest in AI adoption. Nigeria has developed a draft National Artificial Intelligence Strategy, while Kenya recently launched its AI Strategy 2025–2030 and South Africa has unveiled its National AI Policy Framework.
Banks, telecommunications companies and public agencies are also increasingly setting up AI teams and pilot projects.
However, Apampa argued that many of these initiatives risk falling short because the underlying data infrastructure is not sufficiently developed to support AI systems.
He noted that the challenge is not unique to Africa, as many AI projects fail globally due to poor-quality data. But he warned that the consequences could be more severe in African public sector environments, where inaccurate AI-driven decisions could affect access to public services and erode citizens’ trust in government.
“In an African public sector context, the same failure produces wrong decisions about real people, often the people who can least afford to be on the receiving end of a wrong decision,” he said.
The data expert outlined what he described as a five-step process that institutions should complete before implementing AI systems. These include assessing the quality of available data, establishing clear governance structures, embedding controls into data pipelines, maintaining active oversight of data systems, and only then deploying AI models.
He stressed that many organisations are attempting to skip the first four steps and move directly to AI deployment, a strategy he believes could lead to costly failures.
According to him, while flawed data may simply produce misleading reports in traditional analytics systems, AI models built on poor-quality data can amplify errors, automate inaccurate decisions, and make problems more difficult to reverse.
Apampa said the gap between the ambitions contained in many African AI strategies and the current state of the continent’s data infrastructure remains significant, adding that building reliable data systems could take years of sustained investment and governance reforms.
He urged policymakers, academics, and technology practitioners to shift the conversation from the rapid deployment of AI tools to ensuring that the data supporting those tools is accurate, well-governed, and fit for purpose.
“The continent’s AI moment is real, and the energy behind it is right,” he said. “It is just sitting on top of a layer that is not yet ready to carry it.”