AI and Electoral Integrity in Africa: The Battle Against Generative Disinformation

Generative AI technologies are increasingly being used to manipulate elections in Africa, posing significant challenges to democratic integrity. Robust countermeasures, including enhanced digital literacy, fact-checking, and government regulation, are essential to combat AI-driven disinformation and protect democratic processes.

The rapid evolution of generative artificial intelligence (AI) has introduced new challenges to democratic processes in Africa, particularly in the realm of electoral integrity. Research institutes such as the Center for Technology Innovation at The Brookings Institution and Research ICT Africa have been closely examining the implications of AI-generated propaganda on elections. Generative AI technologies, such as deepfakes and AI-generated propaganda, have been increasingly leveraged to manipulate public opinion and disrupt elections across the continent. Notable instances have been observed in countries like Nigeria, Burkina Faso, and Gabon, where AI-generated content has played a role in electoral manipulation and even coups.

AI Propaganda and African Elections

The 2023 presidential election in Nigeria, for instance, saw a flood of AI-generated endorsements and fake audio recordings aimed at swaying voter opinions. Similar tactics were employed during the 2024 elections in Senegal and South Africa, where AI-generated videos and deepfakes were used to spread misinformation about political candidates and their affiliations. These incidents underscore the urgent need for robust countermeasures to protect the integrity of democratic processes. In 2024, over a dozen African countries are scheduled to have elections. Given the existing rate of disinformation that plagues African elections, coupled with social unrest, internet blackouts, and ballot tampering, the increasing ubiquity of generative AI heightens concerns that AI-fueled disinformation campaigns could exacerbate current issues within the African continent.

Fact-Checking: A Crucial Defense

To combat the rising threat of AI-generated disinformation, independent fact-checking organizations across Africa have stepped up their efforts. Groups like Africa Check, PesaCheck, and Dubawa have been pivotal in identifying and debunking false information. These organizations utilize data-driven approaches and collaborate with social media platforms to flag and reduce the spread of misleading content. Despite these efforts, the sheer volume and sophistication of AI-generated disinformation present significant challenges. Social media platforms have enacted various safeguards to hinder the proliferation of disinformation, but many of these efforts are undermined by loopholes such as the lack of moderation for audio content or automated content moderation systems that have limited understanding of low-resource languages. Furthermore, citizens with limited digital and media literacy often inadvertently disseminate propaganda. As efforts increase to curb the spread of AI-generated disinformation, governments must allocate resources toward increasing citizen engagement in AI literacy efforts. Governments will also have to establish regulatory frameworks that can effectively govern the use of generative AI and other emerging technologies while promoting safeguards to prevent the spread of disinformation.

Bridging the Digital Divide

The spread of disinformation is further complicated by the digital divide and varying levels of media literacy among African populations. Many citizens, particularly those in rural areas and older adults, may lack the necessary skills to critically evaluate the information they encounter online. This makes them more susceptible to propaganda and misinformation. To address this, there is a growing call for enhanced digital and AI literacy programs that can equip the general public with the tools to discern credible information from falsehoods. Research has shown that citizens with limited digital and media literacy often inadvertently disseminate propaganda. However, there are instances, such as the 2019 Nigerian elections, where citizens intentionally shared disinformation about specific presidential candidates, knowing this information was false. Research has also shown that AI-generated content is often difficult to detect. While the research is mixed on the specific demographics of populations who are more likely to spread misinformation, older adults and those with lower educational attainment are less likely to identify disinformation. In Nigeria specifically, reports have indicated that younger adults often received WhatsApp broadcasts about viral protection and remedies from older relatives during the 2014 Ebola outbreak. These nuances, combined with the increasing digital divide in Africa, necessitate the development of efforts that ensure marginalized populations such as those in rural areas, older adults, and women are equally involved in educational activities against disinformation.

The Role of Governments and Tech Companies

Governments and tech companies also play a crucial role in this fight. Social media platforms have introduced measures such as limiting message forwarding and employing AI tools to detect disinformation. However, these measures often fall short due to technical limitations and the rapid adaptation of disinformation tactics by malicious actors. Effective regulation and governance are essential to establish standards for AI usage and to hold platforms accountable for the spread of harmful content. The environmental and ethical implications of AI development also cannot be ignored. The training of large AI models consumes significant resources and often involves the exploitation of labor in the Global South. Addressing these issues requires a coordinated effort to ensure that AI technologies are developed and deployed responsibly, with consideration for their broader societal impacts.

Preparing for the Future

As Africa prepares for numerous elections in the coming years, the threat of AI-driven disinformation looms large. Protecting democratic processes will require a multifaceted approach involving government regulation, civil society engagement, and public education. By fostering a just and transparent AI ecosystem, stakeholders can mitigate the risks posed by generative AI and uphold the integrity of elections across the continent.