<i>Responsible AI in Africa: Challenges and Opportunities</i>, edited by Damian Okaibedi Eke, Kutoma Wakunuma, and Simsola Akintoye

With ChatGPT, Dall-E, and other “generative” artificial intelligence tools now widely available, the ethics of synthesizing texts and art have been hotly debated. Universities, libraries, and big tech companies alike have been doing brisk trade in panels, speaker series, and workshops to help us make sense of this new moment. But fears over what AI means for human employment, intellectual property rights, and even the very definition of “human” are not new.In response to these worries—sometimes justifiable, at times hyperbolic—the notion of “responsible computing” has emerged as an ostensible strategy for mitigating the impact of AI technologies on human lives. Broadly speaking, responsible computing encompasses an approach to technological development that foregrounds ethics, social responsibility, transparency, and accountability, with an eye toward human welfare. Responsible AI in Africa: Challenges and Opportunities, a volume edited by Damian Okaibedi Eke, Kutoma Wakunuma, and Simsola Akintoye, is a welcome and much-needed contribution to this space.Despite a common fear that AI will escape the control of its creators, humans are, in fact, the main drivers of the development of these technologies—and that may be scarier than letting the computers be in charge. However unthinkingly, human developers encode their biases into technologies through decisions about the kinds of problems they want to solve; the demographics of populations they aim to serve; and the racial, gender, national, and other stereotypes in the training data for deep learning. The demographics of the tech industry itself include large numbers of white and Asian men but substantially fewer Black, Latinx, and Indigenous people, as well as women. AI developers thus privilege perspectives, ways of knowing, and knowledges from the Global North. Even proponents of responsible computing, well intended as they are, tend to privilege the critiques, analytical frameworks, and humanistic discourses of the Global North—ironically contributing to the marginalization of vital voices from the Global South.Responsible AI in Africa responds to these absences by insisting that developers, governments, policymakers, and even end users—in countries in Africa and abroad—grapple with and address them. The issues raised in the volume are both pressing and not new. The idea of “neural networks” is itself premised on models of human cognition developed by studying white Western European men. Consequently, as I argue in my essay in a recent volume (Bodies of Information: Intersectional Feminism in Digital Humanities, edited by Jacqueline Wernimont and Elizabeth Losh), AI technologies promote a fictive “universal” representation of cognition that is, in fact, particular.Responsible AI in Africa, the first book of its kind, pushes back against the ways that AI developers and adopters—and well-meaning discourses of responsible computing—continually center the Global North as the model, producer, and arbiter of the most powerful technologies today. The volume makes the case that the only responsible approaches to technology are ones that recognize both the colonial histories and contemporary neocolonial practices within tech industries. After all, the Global South—which includes countries in Africa—is a site of disproportionate exploitation for tech industries of the Global North. As the editors and authors in the volume point out, underdeveloped technological infrastructures in most African countries, coupled with financial opportunities through short-term visa programs in the United States, Canada, and Western Europe, promote a “brain drain” that exacerbates uneven technological development. In turn, these factors deprive African countries of access to the capital that a thriving technological sector commands. Likewise, Responsible AI in Africa highlights the neocolonial extraction of data from African countries, and the environmental impacts of African countries serving as sites for the mining of cobalt and the dumping of e-waste. The volume makes a powerful case for recognizing and prioritizing the voices of those who are most disproportionately affected by computing itself.Responsible AI in Africa comprises an introduction and eight chapters that adeptly illuminate these issues. Editors Eke, Wakunuma, and Akintoye provide a thorough and comprehensive look at the landscape of AI development (and underdevelopment) on the African continent. They also outline the key concepts of responsible computing, demonstrating how they apply it to AI in general and in African countries specifically. The eight chapters roughly fall into two categories: broad explorations of AI ethics in African contexts and specific case studies situated in different countries or regions on the continent.While editors and authors are careful to remind readers that African perspectives are manifold, the chapters exploring African AI ethics at times depict the continent as a monolith. This limitation, however, reflects conditions that the volume intends to address in the long term: with the exception of South Africa, Nigeria, Ethiopia, Kenya, and Ghana, most countries in Africa do not yet have thriving technology sectors. The reasons extend back to colonial and postcolonial histories, the drawing of borders by European countries, and the level of neocolonial intervention. What becomes clear in the multiple essays that articulate distinctly “African” approaches to AI is that the values authors ascribe to African communities are consonant with the values of responsible computing. For example, Emma Ruttkamp-Bloem's essay, “Epistemic Just and Dynamic AI Ethics in Africa,” identifies communal values, African notions of personhood, and the relationship between rights and duties in African contexts, and demonstrates their relationship to values of responsible computing in the Global North, such as openness, reproducibility, and sustainability. Chinasa T. Okolo, Kehinde Aruleba, and George Obaido's “Responsible AI in Africa—Challenges and Opportunities” grapples with the ways that countries in Africa are seen as “markets” for exportation of AI technologies, particularly from China. Recognizing that the incursion of foreign technologies is largely inevitable, the authors offer an important vision for local adaptations of imported technologies that could make AI technologies culturally sustaining. At the same time, they make the powerful case for national investment in developing the capacity to produce homegrown AI technologies.The country- and region-specific case studies attest to the importance and significance of local approaches to technological development. They also provide more granular explorations of responsible computing in their national contexts. Favour Borokini, Kutoma Wakunuma, and Simisola Akintoye's essay, “The Use of Gendered Chatbots in Nigeria: Critical Perspectives,” shows how uncritical adoption of technologies developed in the Global North fuels racial and gender stereotypes. Other essays offer important recommendations for integrating AI ethics to mitigate the violence that unchecked technologies can facilitate on the continent. The regional approach proposed in Bernd Carsten Stahl, Tonii Leach, Oluyinka Oyeniji, and George Ogoh's “AI Policy as a Response to AI Ethics? Addressing Ethical Issues in the Development of AI Policies in North Africa” is also notable for its application of an ethics-based approach to North African technological growth. The authors provide a compelling model for how to take more localized approaches to responsible computing. Their intervention is particularly important because Africa is a continent made up of many countries, cultures, ethnicities, languages, and religions, which requires a nuanced approach to technology policy.While the volume includes essays by contributors who work in South Africa, Kenya, and Zambia, more than half of the editors are part of the diaspora—working from the United Kingdom, the Netherlands, Sweden, and the United States. The three editors and seven (out of sixteen) authors all work at De Montfort University in Britain. On one hand, their locations reflect the imperial and neoliberal geopolitics of tech that they are committed to fighting, where employers of the Global North command the resources to lure workers of the Global South away. On the other hand, it raises the question of what responsible AI in Africa would look like if narrated through a broader representation of scholars and practitioners within countries on the African continent and from different regions beyond southern Africa and East Africa.Nonetheless, this is a groundbreaking contribution that clears space for further investigations and clarifies that what African countries need must be driven by thinkers and workers in African countries rather than by the Global North or by the diaspora. The genre of an edited volume foregrounds a plurality of voices and perspectives that sow the seeds for further inquiry on this important topic. The volume also successfully communicates with multiple audiences: scholars and practitioners in African countries and in the Global North, as well as governments and policymakers around the world. Its open access format ensures that readers who are in the position to build on and adapt the useful recommendations within these pages will be able to find them.Responsible AI in Africa provides overdue attention to the importance of designing technological interventions that foreground knowledges, values, and beliefs from African communities. The volume leaves readers with no uncertainty about needed growth areas: infrastructure, governance, the data ecosystem, STEM education, and other factors necessary for the development of AI in African countries. At the same time, its proactive vision provides road maps to overcoming these challenges.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.