Vodacom Warns African Governments Not to Copy the EU AI Act as SA Rewrites Its AI Policy

By

October 21, 2026

Vodacom has urged African governments not to copy the European Union’s AI Act in full. The company warned that heavy European rules could hold back innovation and push out young African developers.

The warning is the main argument of a policy paper Vodacom wrote with law firm Bowmans. The paper, released on Wednesday, is called Governing Artificial Intelligence in Africa: A Principles-based Framework for Responsible Regulation.

“Africa must have a meaningful voice in shaping the rules that will govern its AI future,” Ayman Essam, Vodacom’s group chief officer for external affairs, said at the launch.

The paper comes out while South Africa rewrites its national AI policy. Communications minister Solly Malatsi withdrew the draft in April after fake references were found in it. Malatsi said his department aims to publish a revised version in the first quarter of next year.

According to the paper, four African realities should shape AI rules. These are weak infrastructure, scattered data rules, uneven digital skills and deep inequality. The paper said about 86% of Africa’s workforce is in the informal sector.

The paper builds on six principles. They are transparency, accountability, fairness, safety, privacy and proportionality. Vodacom said these appear in almost every national framework it studied.

The paper also said risk should be judged locally. It gave the example of an AI credit-scoring system. In an economy where most workers are informal, it said, such a system carries a higher risk of shutting people out than in Europe.

Vodacom said it supports the risk-based thinking of the EU AI Act. But the paper said the act depends on testing bodies, market watchdogs and expert regulators that many African countries do not have.

It quoted a European Commission estimate of 6,000 to 7,000 euros in compliance costs per AI application. The commission said that figure applies to an average high-risk system only.

The paper does not mention South Africa’s withdrawn draft. That draft proposed seven new bodies, including a National AI Commission, an AI Ethics Board and a National AI Safety Institute. Vodacom said a standalone AI agency needs resources that most African countries cannot provide right now.

Instead, the paper recommends one lead body inside an existing ministry or regulator. Sector regulators would enforce AI rules in their own areas. An advisory council with different groups would also give input.

The Department of Communications appears to be moving in a similar direction. In July, acting deputy director-general Jeanette Morwane told TechCentral that each sector would write its own AI strategy based on national guidelines.

“We don’t necessarily need to create a plethora of new institutions,” Khusela Diko, chair of parliament’s communications portfolio committee, said at the time. She added: “It needs to go beyond principles; we understand the principles.”

The paper sets out a plan in three phases. First come a national AI strategy, a lead body and basic openness rules for high-risk systems. These include systems that decide access to credit, healthcare or jobs. Later phases add impact checks, incident reports, independent audits and registers of high-risk systems.

Vodacom admitted it has an interest in the result. The company said that as both a user of AI and a supporter of reform, its “dual role” calls for openness. Under its model, AI in telecoms would fall under the telecoms regulators Vodacom already deals with.

Bowmans’ Heather Irvine also sits on the independent expert panel Malatsi appointed in May to help rebuild South Africa’s AI policy.

The paper said regional groups such as the Southern African Development Community may be a more realistic route than binding continental treaties. It said the African Union’s Malabo Convention on cybersecurity and data protection has been ratified by only 20 of its 55 member states.

Source: this article is based on reporting by TechCentral. Image: TechCentral. Written with the help of AI and published by the Tzaneen Voice Technology desk. See our Editorial Standards.