As artificial intelligence (AI) tools become more embedded in African education, experts are raising concerns that these technologies often carry the cultural values of their creators – typically from the Global North. While AI promises personalized learning and improved classroom efficiency, its design may inadvertently promote Western ideologies, suppress Indigenous knowledge, and reshape local education systems in ways that clash with African contexts.

This critical perspective, recently featured on the UNESCO website, highlights how educational AI tools like ChatGPT and Gemini often overlook Africa’s unique ecological, spiritual, and communal learning systems. For example, when asked about seasonal cycles, these tools default to Western norms—such as four seasons—ignoring the two-season reality in much of West Africa. Similarly, the communal learning practices of groups like the Maasai and Kipsigis in Kenya are at odds with the individualized, screen-based learning AI often encourages.

Multilingual as they may be, most AI tools remain monocultural, lacking the depth to represent the epistemologies embedded in African languages. This leads to what scholars call “digital-territorial coloniality,” where imported technologies subtly displace local worldviews.

Grassroots efforts like the Masakhane project are actively pushing back. By building AI tools in African languages through participatory research, the initiative empowers local communities to reclaim control over their digital futures. Masakhane’s work includes translating STEM research into African languages and creating language models trained on local content.

Without it, Africa risks reinforcing technological dependence instead of shaping its own educational and cultural destiny.