2025 State of Internet Freedom In Africa Report Documents the Implications of AI on Digital Democracy in Africa
By Juliet Nanfuka |
The 2025 edition of the Forum on Internet Freedom in Africa (FIFAfrica25) concluded on a high note with the unveiling of the latest State of Internet Freedom in Africa (SIFA) report. Titled Navigating the Implications of AI on Digital Democracy in Africa, this landmark study unpacks how artificial intelligence is shaping, disrupting, and reimagining civic space and digital rights across the continent.
Drawing on research from 14 countries (Cameroon, Egypt, Ethiopia, Ghana, Kenya, Mozambique, Namibia, Nigeria, Rwanda, Senegal, South Africa, Tunisia, Uganda, and Zimbabwe), the report documents both the immense promise and the urgent perils of AI in Africa. It highlights AI’s potential to strengthen democratic participation, improve public services, and drive innovation, while also warning of its role in amplifying surveillance, disinformation, and exclusion.
Using a qualitative approach, including literature review and key informant interviews, the report shows that AI is rapidly transforming how Africans interact with technology, yet AI also amplifies existing vulnerabilities, introduces new challenges that undermine fundamental freedoms, and deepens existing inequalities.
The report notes that the political environment is a crucial determinant of AI’s trajectory, with strong democracies generally enabling a positive outcome. Top performers in freedom and governance indices such as South Africa, Ghana, Namibia, and Senegal are more likely to set the standard to AI rollout in Africa. Conversely, countries with lower democratic credentials such as Cameroon, Egypt, Ethiopia, and Rwanda risk constraining AI’s potential or deploying it to amplify digital authoritarianism and political repression.
Countries such as South Africa, Tunisia and Egypt that have a higher internet access and technological development, Gross Domestic Product (GDP) per capita, and score highly on the Human Development Index (HDI), are more likely to lead in AI. Meanwhile, countries with lower or weaker levels of digital infrastructure face greater challenges and higher risks of AI replicating and worsening existing divides. Such countries include Cameroon, Mozambique and Uganda.
The political environment is a crucial determinant of AI’s trajectory, with strong democracies generally enabling a positive outcome. Economic and developmental status also dictates the capacity for AI development and adoption.
Despite these challenges, the report documents that AI offers substantial value to the public sector by improving service delivery and enhancing transparency. Governments are leveraging AI tools for efficiency, such as the South African Revenue Services (SARS) AI Assistant for tax assessments and Nigeria’s Service-Wise GPT for streamlined governance document access. In Kenya, the Sauti ya Bajeti (Voice of the Budget) platform fosters fiscal transparency by allowing citizens to query and track government expenditures. Furthermore, countries like Tunisia and Uganda are using AI models within tax bodies to detect fraud, while Rwanda is deploying AI for judicial system improvements and identity management at borders.
The private sector and academic institutions are driving AI-inspired innovation, particularly in the areas of FinTech, AgriTech, and Natural Language Processing (NLP). For the latter, notable efforts to localise AI include Tunisia’s TUNBERT model for Tunisian Arabic, and Ghana’s Khaya, an open-source AI-powered translator tailored for local languages. In Ghana, the DeafCanTalk, is an AI-powered app that enables bidirectional translation between sign language and spoken language, and has enhanced accessibility for deaf users. Rwanda has integrated AI into healthcare using drone delivery systems for medical supplies, while Cameroon and Uganda use AI to assist farmers with pest identification.
However, despite increasing investment, such as the ongoing USD 720 million investment in compute power by Cassava Technologies across hubs in South Africa, Egypt, Kenya, Morocco, and Nigeria, Africa receives significantly lower AI funding than global counterparts.
Moreover, while AI is gaining traction across many sectors, the proliferation of AI-generated misinformation and disinformation is a pervasive and growing challenge that poses a critical threat to electoral integrity. During South Africa’s 2024 elections, deepfake videos were circulated to manipulate perceptions and endorse political entities. Similarly, during elections and protests in Kenya and Namibia, deepfake technology and automated campaigns were used to discredit opponents.
The report also documents that governments are deploying AI-powered surveillance technologies, which has led to widespread privacy violations and a chilling effect on freedoms. For example, pro-government propagandists in Rwanda utilised Large Language Models (LLMs) to mass-produce synthetic messages on social media, simulating authentic support and suppressing dissenting voices. Meanwhile, algorithmic bias and exclusion are producing discriminatory outcomes, particularly against low-resource African languages. Also, AI-based content moderation is often ineffective because it lacks contextual understanding and fails to capture local nuance.
A key finding in the report is that across the continent, the pace of AI development far outstrips regulatory readiness. None of the 14 study countries has AI-specific legislation. Instead, fragmented laws on data protection, cybercrime, and copyright are stretched to cover AI, but remain inadequate. Data protection authorities are under-resourced, under-staffed, and often lack the technical expertise required to audit or govern complex AI systems.
Although many national AI strategies are emerging, they prioritise economic growth while neglecting human rights and accountability. This is also fuelled by policy processes that are often opaque and dominated by state actors, with limited multistakeholder participation.
The report stresses that without deliberate, inclusive, and rights-centred governance, AI risks entrenching authoritarianism and exacerbating inequalities.
To avoid the current trajectory that AI is taking in Africa, in which AI risks entrenching authoritarianism and exacerbating inequalities, the report calls for a human-centred AI governance framework built on inclusivity, transparency, and context.
It also makes recommendations, including enacting comprehensive AI-specific legislation, instituting mandatory human rights impact assessments, establishing empowered AI and data governance institutions, and promoting rights-based advocacy. Others are building technical capacity across governments, civil society and media, and developing policies that prioritise equity and human dignity alongside innovation.
AI offers Africa the opportunity to foster innovation, strengthen democracy, and drive sustainable development. This edition of the State of Internet Freedom in Africa report provides an evidence-based roadmap to ensure that Africa’s digital future remains open, inclusive, and rights-respecting.Find the report here.
