Vodacom proposes human rights-based framework for African AI governance

A unified approach for Africa

Vodacom, a major telecommunications company, recently unveiled a policy paper titled 'Governing AI in Africa: A Principles-based Framework for Responsible Regulation.' The document was created in partnership with the law firm Bowmans to address the unique challenges of deploying artificial intelligence across the African continent.

One of the primary hurdles identified by Vodacom is regulatory fragmentation. Currently, a company looking to deploy an AI tool in a single country might have to engage with several different regulators. For example, the company noted that in Tanzania, a firm might need to deal with three regulators, while in Mozambique, there are two.

The inconsistency in these processes creates significant friction. Some regulators may provide answers within 10 days, while others can take as long as six months. This variability makes it difficult for companies to maintain project timelines or scale AI technologies effectively across different borders.

Prioritizing human rights

Rather than simply copying existing laws from other regions, the paper advocates for a framework rooted in human rights. Vodacom's acting managing executive for public policy and regulatory affairs, Mosa Thekiso, stated that the goal is to promote a 'human rights-based baseline' for the continent.

The framework highlights six essential governance principles: transparency and explainability, accountability and human oversight, fairness and non-discrimination, safety and security, privacy and data protection, and proportionality. These principles are intended to ensure that AI systems do not infringe on individual freedoms or dignity.

The paper suggests a 'horizontal' regulatory model. This means establishing broad, overarching rules that apply across the board, with specific, sector-based rules only being introduced where they are genuinely needed. This approach is designed to be proportionate, meaning the level of oversight should be phased based on the specific risk of the technology and the readiness of the country involved.

Learning from global models

To build this framework, the authors looked at various international jurisdictions to see what worked and what did not. The paper specifically examined the regulatory paths taken by the European Union, China, and Singapore. By analyzing these different models, the authors aim to avoid past mistakes while tailoring the rules to the specific context of Africa.

Ayman Essam, Vodacom Group's external affairs chief, emphasized that transparency is vital because AI learns from data sets that are chosen by humans. He argued that consumers deserve to know what data is being fed into machines that eventually influence life-altering decisions in banking, healthcare, and education.

Without proper guardrails, Essam warned that AI could reach incorrect or even life-threatening conclusions. He used the example of a banking algorithm potentially blocking funds based on a pattern of overspending, illustrating how automated decisions can have immediate, real-world consequences for individuals.

A unique opportunity for the Global South

The partnership highlights a unique historical moment for the African continent. Bowmans partner Kate Paterson noted that Africa has never had an opportunity like this before, comparing the current AI shift to the industrial revolution. Because Africa is a major part of the Global South, it has a chance to determine how it wants to be involved in the AI era from the very beginning.

The paper suggests that the most appropriate response is for regions to collectively decide how to negotiate the future of AI. This collaborative approach aims to protect human rights from the outset, ensuring that the technology serves the needs of the people rather than just being adopted without oversight. By establishing these principles now, the goal is to create a balanced environment where innovation can flourish without compromising safety.