Africa’s role in generating indigenous content to shape AI narrative, address algorithm bias
By Omoruyi Edoigiawerie Esq
Last week, I attended the Global Inclusivity and AI: Africa (GIAA) Conference jointly organized by the Nigerian and United States governments, where engaging with stakeholders and participating in discussion sessions brought to light several critical issues surrounding artificial intelligence (AI) and its application in developing and underdeveloped nations.
These insightful exchanges are not only expedient at this time, but they are also bringing to the fore the need for Africa to play a role in influencing the AI narrative and addressing algorithmic bias.
AI has become one of the most transformative technologies of the 21st century, permeating every aspect of our lives, from healthcare to finance, education, and entertainment. However, like most technologies, AI is not neutral.
It reflects the values, priorities, and biases of those who create it. As AI algorithms increasingly dictate what we see, hear, and experience, they also have the potential to reinforce and perpetuate biases, particularly against underrepresented groups.
This is where Africa, with its rich cultural diversity, untapped intellectual resources, and unique perspectives, can play a pivotal role in shaping the AI narrative and correcting algorithmic biases.
*Understanding algorithmic bias
Algorithmic bias refers to the systematic and unfair discrimination encoded into AI systems. These biases arise when AI models are trained on skewed or incomplete datasets reflecting societal prejudices. When the majority of AI training data comes from Western or more developed nations, these systems tend to overlook or misinterpret the experiences and nuances of marginalized groups, including those in Africa.
For example, facial recognition software is less accurate in identifying people of African descent due to a lack of representation in the training datasets. Similarly, language models, at the core of many AI-driven applications, often fail to accurately understand or generate African languages. This has far-reaching consequences, from perpetuating stereotypes to denying critical services.
AI bias, therefore, is not merely a technical issue; it is also a socio-political one. It affects the daily lives of billions of people and potentially exacerbates existing inequalities.
While many Africans are fluent in their native languages, a significant gap exists regarding literacy in these languages. Although oral communication remains strong across the continent, the ability to read and write in indigenous languages is alarmingly low. This disparity is primarily due to colonial legacies that prioritized foreign languages, such as English, French, and Portuguese, in formal education systems. As a result, many Africans are proficient in their spoken mother tongues but often lack the skills or formal training to read or write. This has led to a situation where rich cultural knowledge is predominantly passed down orally, leaving indigenous languages underrepresented in written forms, digital content, and technological platforms, which are increasingly dominated by Western languages.
READ ALSO: Advisory boards: Secret weapon Nigerian start-ups need now
Africa’s underrepresentation in AI datasets contributes to its being left out of critical technological advancements. But this challenge also presents a unique opportunity: Africa can lead the development of AI technologies that reflect its values, cultures, and priorities, ultimately addressing algorithmic bias at its root.
*Africa’s rich indigenous knowledge and cultural diversity
Africa has over 1.3 billion people and boasts more than 2,000 distinct languages and cultures. Its cultural diversity, oral traditions, indigenous knowledge systems, and unique problem-solving methods are invaluable resources that can contribute to the global AI narrative. Yet, much of this indigenous knowledge is underrepresented in the digital world and, by extension, in the datasets that power AI algorithms.
To tackle AI bias, Africa needs to generate and curate indigenous content that reflects the continent’s diversity. This includes digitizing African languages, folklore, history, music, and art and documenting traditional practices in agriculture, medicine, and conflict resolution. By integrating these into AI systems, Africa can help create more inclusive and context-aware technologies that respect and understand the continent’s cultural richness.
*The importance of African-led AI research
One of the main reasons AI biases persist is the underrepresentation of African researchers in AI development. AI research is often dominated by institutions and companies in the Global North, which means that the challenges and realities faced by Africans are rarely prioritized. However, several African researchers and institutions are starting to take the lead in AI innovation, focusing on issues most relevant to the continent.
African researchers are uniquely positioned to challenge existing AI paradigms and contribute indigenous perspectives that have traditionally been overlooked. For example, African linguists and computer scientists can develop natural language processing (NLP) models for African languages, making AI applications accessible to millions of people who speak languages other than English, French, or Arabic. By embedding African languages into AI systems, we can ensure that the continent’s voices are heard and understood by global technologies.
Moreover, African-led AI research can focus on issues uniquely relevant to the continent, such as sustainable agriculture, disease prevention, and infrastructure development. For instance, AI algorithms can be tailored to predict crop yields in Africa’s diverse climates or to detect diseases like malaria more accurately. When Africans lead AI research, they can ensure that the technology serves the continent’s needs rather than merely replicating Western priorities.
*Data sovereignty and ethical AI development in Africa
One of the critical challenges in addressing algorithmic bias is data sovereignty. Many African countries are still developing legal frameworks for data protection, which means that large tech companies often extract data from the continent without adequately considering the privacy and security of African citizens. This “data colonialism” seriously impacts how AI technologies are developed and deployed.
To shape the AI narrative, Africa must take control of its data. This means developing policies that protect data sovereignty and ensure that African data is used ethically and for the continent’s benefit. African governments need to collaborate with tech companies and research institutions to establish frameworks for responsible AI development that promote African culture.
At the same time, ethical AI development requires transparency and accountability. African institutions and researchers should push for open-source AI models and datasets that are transparent and auditable. This will not only foster trust in AI technologies but also allow African stakeholders to challenge biases and ensure that AI serves the continent’s interests.
*AI in education: Cultivating African talent
To generate enough indigenous content and correct algorithmic bias, Africa needs to encourage indigenous AI practitioners who understand the continent’s realities. Education is key to this transformation. African universities, governments, and private institutions should invest in AI education and research, ensuring students have the skills to contribute to AI innovation.
While programs currently exist that are centred on building an AI ecosystem on the continent. However, more must be done to expand access to AI education, particularly in underserved regions. By training in AI, machine learning, and data science, African institutions can ensure that the next generation of tech leaders is well-prepared to challenge AI biases and create technologies that benefit their communities.
Moreover, African AI talent must be nurtured to lead the development of indigenous content. Students and researchers can be encouraged to work on AI projects that document and digitize African languages, oral traditions, and indigenous knowledge. By creating AI models that reflect the continent’s unique perspectives, African innovators can ensure that global AI systems are more inclusive and equitable.
*The role of African governments and the private sector
African governments have a crucial role in fostering AI innovation and ensuring that the technology benefits all citizens. By investing in digital infrastructure, supporting AI research, and creating favourable policies for AI start-ups, African governments can spur the development of AI technologies that address local challenges and contribute to global advancements.
In addition to government support, the private sector must play a leading role in funding AI research and development, and this is why Google’s 100 million Naira fund for 10 AI-driven start-ups is a step in the right direction – one which other private sector players should emulate. African start-ups, in collaboration with international companies, can generate the indigenous content needed to correct algorithmic bias and develop AI solutions that are culturally relevant. Initiatives like Google’s AI Research Centre in Ghana and Meta’s investment in African tech hubs demonstrate the growing interest in Africa’s potential to shape the AI narrative.
However, these partnerships must be mutually beneficial, with African stakeholders driving the conversation. Africa cannot afford to be a passive participant in AI development. Instead, it must be an active contributor, shaping the direction of the technology to reflect its values and priorities.
*Africa as a key player in the global AI narrative
As AI continues to influence every aspect of modern life, Africa’s role in generating indigenous content and addressing algorithmic bias is more critical than ever. By leveraging its rich cultural diversity, fostering AI research, protecting data sovereignty, and investing in AI education, Africa can ensure its voice is heard in the global AI conversation.
The task is not just about correcting biases but also about shaping AI to serve the continent’s unique needs and aspirations. In doing so, Africa can help build a more inclusive and equitable technological future that respects and reflects the diversity of human experience. The global AI narrative cannot be complete without Africa, and the time to act is now.
Omoruyi Edoigiawerie is the Founder and Lead Partner at Edoigiawerie & Company LP, a full-service law firm offering bespoke legal services focusing on start-ups, established businesses, and upscale private clients in Nigeria. The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances. His firm can be reached by email at [email protected].
