African Policy Paper · Analysis note no. 01 · October 2026

AI Governance in Africa: From Technological Dependency to Strategic Sovereignty

How can African states regulate critical algorithms, preserve their data heritage and turn artificial intelligence into an instrument of public power?

Author
SOADAN Koffi Sylvain
Theme
AI Governance & Sovereignty
Affiliation
IAI-Togo · University of Lomé
Type
Policy note

Executive summary

The European Union favours a risk-based approach (EU AI Act) while other powers pursue industrial domination strategies. African nations risk being reduced to consumers of exogenous technologies. To avoid a new form of digital extraversion, Africa must build proactive governance based on data sovereignty, model auditability and the training of a multidisciplinary elite able to arbitrate technological choices.

Key recommendations

  1. Create an AI digital passport: make registration mandatory for any AI system deployed in African public services (health, justice, civil registry, taxation).
  2. Establish regional ethics assessment centres: pool algorithmic audit capacity within regional economic communities (ECOWAS, EAC, SADC).
  3. Enshrine the non-export principle for health and identity data: condition model training on local hosting and the development of local skills.

1. The trap of algorithmic extraversion

The growing use of foreign-developed models in African administrations introduces silent biases. Models trained on exogenous legal or medical data reflect neither local realities, nor tropical epidemiology, nor African languages.

2. From the Malabo Convention to the African Union AI Strategy

The adoption of the continental AI strategy by the African Union in 2024 marks a turning point. A political text without technical infrastructure, however, remains inoperative: regulators must be equipped with software inspection tools able to analyse the code, weights and telemetry of AI systems.

3. Conclusion

Africa’s technological future will also be decided in the multilateral forums where global standards are negotiated (ITU, UNESCO, UN). Training professionals able to engage with engineers, diplomats and judges is a condition of future sovereignty.

How to cite this publication

SOADAN Koffi Sylvain (2026). “AI Governance in Africa: From Technological Dependency to Strategic Sovereignty”, SKS Policy Notes & Analysis, Note no. 01, Lomé, Togo.

Back to research