In Accra, the question is no longer only how artificial intelligence will transform Africa, but who will set its rules and who will actually be able to benefit. The African Internet Governance Forum (AfIGF), running from November 2 to 5, devotes several sessions this Tuesday, November 3, to AI governance, connectivity, digital rights and youth participation. The agenda is not an agreement already reached: it opens a space for discussion whose impact will depend on what is translated into policies, budgets and practice.
That distinction matters. Digital conferences often produce ambitious declarations while the most immediate obstacles remain part of everyday life: the price of data, network quality, access to devices, protection of personal information and public administrations’ ability to apply consistent rules. The Accra gathering puts these issues into one conversation, bringing together public authorities, technical communities, companies, researchers, civil society and young participants.
A Day Connecting AI with Rights and Inclusion
The official November 3 programme includes a discussion on a pan-African approach to AI governance, a session on digital sovereignty and another on child safety in the age of AI. Other exchanges address internet rights and freedoms, AI use for rural women in Uganda and the place of young people in digital institutions. The choice points to a central idea: governance is not only about AI models or regulations. It also concerns services, languages, freedoms and people who remain on the margins.
The Forum is part of a broader continental effort. The African Union’s Digital Transformation Strategy for 2020–2030 links digital technology to fields as varied as education, health, agriculture, public services and trade. The challenge now is to connect this vision to national frameworks: rules that protect citizens without freezing innovation, and cooperation mechanisms that prevent every country from having to start from scratch.
Sovereignty Does Not Mean Isolation
In the AI debate, digital sovereignty can mean control over data, the ability to choose infrastructure, command of skills and the power to negotiate with major platforms. But a sovereignty policy is not simply a matter of closing digital borders. Companies, researchers and public administrations need compatible standards, interoperable services and cross-border partnerships. The phrase “sovereignty without isolation,” which appears in the programme, captures this productive tension: increasing local decision-making capacity while preserving the ability to cooperate.
That cooperation is particularly important for AI. Automated systems can be trained on data that poorly reflect local languages, practices and realities. Access to quality datasets, people’s consent, data protection and the ability to challenge an automated decision are therefore economic as well as legal questions. A company unable to explain its decisions or secure customer data risks losing trust, even if its tool appears powerful.
Connectivity Remains the Most Immediate Test
Figures from the International Telecommunication Union underline the gap between promise and use. In its 2025 regional report, the ITU estimates that 38% of Africa’s population used the internet in 2024, compared with 68% worldwide. Mobile broadband coverage reached 86% of the population, but coverage does not automatically mean that a service is affordable, stable or fast enough. The ITU also reports a marked divide between urban and rural areas: usage stood at 57% in cities versus 23% in rural communities.
These figures change how an AI strategy should be assessed. An automated agricultural service may look promising on paper, but it must work with the phones and networks available, in languages its users understand and at a realistic cost. An administration may announce a digital portal; if a significant share of the public cannot connect or does not know how to protect its login details, the service is not truly universal. “Meaningful connectivity” therefore requires quality, affordability, skills and security, not simply an antenna on a map.
Measurable Outcomes, Not Another Vocabulary
For policymakers, the value of a forum such as AfIGF will be measured less by the number of panels than by what follows them. Practical targets could include accessible reporting mechanisms, understandable rules on data use, child-appropriate safety assessments, or common standards that facilitate cross-border services. Commitments must also be verifiable: who is responsible, what timeline is set, what resources are planned, and how can affected communities hold decision-makers accountable?
The presence of sessions devoted to young people, digital rights and rural communities is a reminder that participation must not be tokenistic. People affected by technology often know about failures its designers do not see: a missing local language, data costs, bias in a service, exposure risks or a lack of appeal. Including them early in design and evaluation can build trust and prevent innovation from deepening the inequalities it claims to reduce.
Accra’s Real Challenge
AfIGF 2026 cannot by itself resolve the infrastructure, financing and skills gaps across the continent. It can, however, help align priorities and build more coherent rules. The November 3 programme places AI alongside rights, safety and connectivity: a combination that avoids treating technology as separate from the material conditions around it.
The best measure of success will not be the abstract promise of “African AI.” It will be the ability to build systems that are useful, challengeable, safe and accessible, while giving countries and communities room to choose their own path. In Accra, the debate is open. The decisive work will begin when principles discussed in conference rooms have to be translated into services, budgets and daily life.
