AI and the Future of Democracy in Africa: Navigating the Promise and the Peril

Research ICT Africa’s recent study, How Might AI Reshape Democracy on the African Continent, wrestles with the high-stakes question: how could AI empower or imperil democracy across the continent? Our analysis reveals no definitive answers, but rather circumstantial possibilities and tensions.

Part of the recent history of the continent is marked by struggles for democratic self-determination against the legacy of colonial underdevelopment, exploitation and oppression. The struggles for freedom and development are complex. Democracy has been slow to take hold in Africa, and many elections have been marred by fraud, violence, and intimidation. AI could either enhance or endanger this struggle, depending on how it is designed, deployed and used.

Concurrently, while some Africans are thoughtfully considering how to harness AI’s possibilities, the technology’s emergence on the continent is largely being shaped by more powerful external actors who often neglect – or are indifferent to – African needs and values. Major technology companies and governments based outside Africa continue developing and deploying AI systems with little regard for development and democracy on the continent. In fact, certain applications of AI, like mass surveillance technologies, are being actively marketed and exported in ways that directly undermine human rights and democratic participation within African nations.

Even so, the rhetoric surrounding AI often presumes African societies have the full autonomy to determine AI’s role. In reality, many impactful AI systems are imposed on African societies via global technology power structures and business interests that are adjacent to principles of democratic accountability. Unless African proponents of democracy can gain more influence over AI, the technology risks amplifying externally driven exploitation rather than locally-determined priorities. Genuine African advancement requires not just wrestling with AI’s practical possibilities, but also contesting the global power relations shaping its emergence on the ground.

The Role of African Proponents of Democracy

The purpose of this report is to give African proponents of democracy a contextual and political understanding of AI and related technology. By African proponents of democracy, which we refer to in the report as African democrats, we mean advocates for universal enfranchisement, self-determination, political participation and human rights. These advocates may be working in government, civil society organizations and institutions, as well as in industry. Our report aims to provide insights for African democrats that are not directly working in the fields of science, technology and innovation policy.

Any analysis of how the deployment of AI systems could impact African democracy must carefully consider the nature of the specific AI technologies involved as well as their unique capabilities and limitations. This includes the surrounding institutional and social contexts, existing policy and regulatory frameworks, plus global dynamics around business and governance. For instance, classificatory and ranking algorithms used by social media platforms to order users’ news-feeds can profoundly shape the distribution and circulation of both civic information and misinformation.

Meanwhile, the emergence of generative AI systems that can synthesize fake images, videos and texts threatens to allow the mass production of fabricated content that may be used to manipulate public discourse. These basic examples illustrate how the impacts of AI on democracy emerge from the interactions between the technical particulars of AI systems, their application environments and wider socio-political forces. Analysis of what AI is and how the deployment of these systems could impact the exercise of African democracy must also consider the institutional structures, the policy space and regulatory environment, as well as global market dynamics, as each of these comes to shape how AI is deployed and used in any particular setting.

Research ICT Africa aims to meaningfully contribute to this larger project by focusing on specific African situations to examine how democratic and advanced technological forces constitute or contradict one another. To effectively deal with these forms, policymakers and practitioners concurrently require comparative knowledge, good practice examples, peer dialogue, as well as analysis and advice tailorable to their contexts. We have sought to provide some insights for this initiative, knowing that considerably more research is required for policymakers and practitioners to have a comprehensive understanding of how AI influences political life in Africa.

Our approach is focused on the political sociology and legal aspects of AI. Even as the research from the sociological study of AI is in its early stages, extrapolation from previous rounds of technological diffusion suggests that AI will reconstitute rather than overhaul existing political structures.

Unmet Expectations

When evaluating the state of democracy in Africa, it is important to keep in mind issues related to state structure and unmet citizen expectations. Many African states face challenges of weak institutional capacity, corruption and neopatrimonialism, which undermine democratic accountability. Furthermore, citizens across Africa often harbor unmet expectations for economic development, service delivery, and political reform, which has led to disillusionment. This means there are fewer people eager to fight for democracy in the face of the risks presented by AI.

Our overall analysis suggests that AI is not yet a primary threat to the integrity of the democratic processes, but it could become one soon if left unchecked given that AI has the power to subvert the core values of democracy, such as participation, accountability, and transparency while creating new mechanisms for manipulation, deception, and coercion. We therefore urge caution in the face of this emerging challenge, which could have profound implications for the future of democracy in Africa.

The purported benefits of AI to democracy can at times be overstated, as users’ experiences are shaped by profit-driven platforms deploying opaque algorithms. In elections, online discourse is further distorted by well-resourced actors employing advanced digital strategies. A case in point is Kenya’s now famous slogan ‘Vitu Kwa Ground’ (roughly, ‘things are different on the ground’) which warns that a candidate’s online popularity may not translate into physical ballots. This phrase encapsulates a recognition that despite the proliferation of social media, on-the-ground organizational strength remains pivotal in mobilizing voters. While AI – and the online sphere in general – presents new opportunities for political engagement, its role in democratization depends on context, platform architectures driven by commercial imperatives, and power dynamics offline.

Decisions about regulation are key to the future of AI in Africa. Without seizing the current regulatory window of opportunity given the novelty of the technology, even if the odds seem unfavorable, AI will likely widen the gap between the rich and the poor, the urban and the rural, and the educated and the illiterate, thereby reinforcing the digital divide.

Emerging Trends for African Democracy

From our vantage point, we see current discourse about AI in Africa orbiting two poles: excessive skepticism and excessive celebration. Some of these tropes can be seen in the various discourses emphasizing economic opportunity, ethical concerns, and developmental aspirations, typically through the aspirational lens of technological nationalism. By providing a tangible point of departure with the democratic process, our aim is to add more dimensions to the present discussion of AI in Africa, so that ultimately more sociologically-derived considerations can be tabled in discussions, assessments, and decisions. In the final chapter of the report, we also suggest avenues of inquiry for coming rounds of policy research in these and adjacent areas.

Our report also identifies and discusses selected emerging trends around AI and African democracy, focusing on the implications of AI on election integrity for democratic elections. AI systems have the potential to transform the political landscape in Africa. As citizens and researchers have observed with social media and online search platforms before, these systems could enable new forms of political participation and contention, but also new forms of surveillance and control. These developments pose challenges and risks to election integrity.

One of the main challenges is the lack of significant regulatory coordination at the global level to subordinate AI product development and commercial uses to the public interest. This has led to massive privacy violations, data manipulation, misinformation and disinformation, cyberattacks, social polarization, and human rights abuses. For example, AI-powered facial recognition systems are being used to surveil and suppress political opponents and activists, as is currently the case in Uganda and may soon be in Zimbabwe.

A related question around AI and democratic backsliding involves how African political parties are funded and whether they can afford to purchase and use AI systems for their campaigns. AI systems could provide advantages for parties that have access to them, such as targeting voters, mobilizing supporters, and influencing public opinion. However, they could also create inequalities, distortions and unfair advantages in the democratic process along the lines of gender, class, and political party affiliation.

Recognizing that technologies are encoded with values and put in service of political and social projects, the use of AI poses a significant risk to pro-democracy movements in Africa. And so African democrats must work to subordinate AI products to the democratic will so they do not impinge upon the rights and dignity of the African people.

Source Link
Featured Image: Image by Alan Warburton / © BBC / Better Images of AI / Quantified Human / CC-BY 4.0