Follow Us on Google
INTRODUCTION
The last installment of this intervention traced the evolution of AI, reviewed notable developments in its trajectory, its African dimension and policy trend therein and beyond. This week’s feature goes further afield, reviewing the position in the US, the EU and China. Thereafter, we consider the dangers of weak localisation and disproportionate influence of foreign technology on African innovation ecosystem. This is followed by a discussion of the issues generated by AI policy and what African states need to do – using Nigeria as an example/template. Enjoy.
USA, EU, CHINA’S PREFERENCES (Continues)
In Africa, the policy landscape is accelerating but uneven. The Global AI Index (www.diplomacy.edu/resource/report-stronger-digital-voices-from-africa/ai-africa-national-policies/ > (Diplomacy.Edu) Accessed on 10th September, 2025) categorizes most African countries as lagging: Egypt, Nigeria and Kenya as nascent, and Morocco, South Africa and Tunisia as waking up (Techpoint Africa, < www.facebook.com/TechpointAfrica/posts/africas-ai-policy-why-a-copy-and-paste-approach-will-fail-this-time-every-countr/1064672189125910/> (Facebook.com, 22nd July, 2025) Accessed on 10th September, 2025). Mauritius led with an AI strategy (Mauritius Artificial Intelligence Strategy, November, 2018 < https://treasury.govmu.org/Documents/Strategies/Mauritius%20AI%20Strategy.pdf > (Treasury.govmu.org) Accessed on 10th September, 2025), followed by Kenya’s AI and blockchain task force (2019) (Kenya Artificial Intelligence Strategy < https://ict.go.ke/sites/default/files/2025-03/Kenya%20AI%20Strategy%202025%20-%202030.pdf > (Ict.go.ke) Accessed on 10th September, 2025), its Digital Master Plan (2022) (Kenya Digital Master Plan, 2022 – 2032 < https://cms.icta.go.ke/sites/default/files/2022-04/Kenya%20Digital%20Masterplan%202022-2032%20Online%20Version.pdf > (Ict.go.ke) Accessed on 10th September, 2025), and Rwanda’s AI policy (Thompson Gyedu Kwarkye, ‘AI policies in Africa: lessons from Ghana and Rwanda’ (TheConversation.com, 25th April, 2025) < https://theconversation.com/ai-policies-in-africa-lessons-from-ghana-and-rwanda-253642 > Accessed on 10th September, 2025), which reflects its national security priorities. Nigeria, Ghana, Uganda, Algeria and South Africa have also announced or drafted
AI policies, often framed around economic growth and innovation.
Continental initiatives, such as the African Union’s Digital Transformation Strategy (African Union, ‘THE DIGITAL TRANSFORMATION STRATEGY FOR AFRICA (2020-2030)’ < https://au.int/sites/default/files/documents/38507-doc-dts-english.pdf > Accessed on 10th September, 2025) and the World Bank’s DE4A program (< www.worldbank.org/en/programs/all-africa-digital-transformation > Accessed on 10th September, 2025), emphasize infrastructure, skills and inclusion, but implementation remains fragmented.
Still, foreign influence looms large. Many African AI and data governance frameworks are modeled directly on external templates, particularly the EU’s General Data Protection Regulation (GDPR) (< https://gdpr.eu/what-is-gdpr/ > Accessed on 10th September, 2025). Nigeria’s NDPR (< https://nitda.gov.ng/wp-content/uploads/2021/01/NDPR-Implementation-Framework.pdf > Accessed on 10th September, 2025), a near copy of the GDPR, introduced concepts like consent, data subject rights and cross-border transfers. While it helped raise awareness and created local compliance industries, it omitted key protections (such as breach notifications, children’s rights and strong enforcement). Similar GDPR-inspired laws have been enacted in Ghana, Kenya and South Africa. This copy-paste strategy provides structure but often lacks localization, leaving gaps in enforcement and contextual fit (Bolu Abiodun ‘Africa’s AI policy: Why a copy and paste approach will fail this time’ (Techpoint.Africa, 22nd July, 2025) < https://techpoint.africa/insight/africas-ai-policy-copy-paste/ > Accessed on 10th September, 2025).
Critics warn that the real problem is not copying but exclusion. As Mozilla’s Kiito Shilongo and other researchers argue, many African AI policies are drafted with heavy input from foreign agencies and consultants, while local communities, startups, and civil society are sidelined. This participatory deficit means policies risk reflecting donor interests more than citizens’ rights. In Rwanda, for example, AI policy was shaped through government agencies and international NGOs with a strong focus on security. Ghana’s was more inclusive, involving startups, academia and telecoms, but leaned toward development goals over safety. Both approaches highlight the political nature of AI policymaking and the different ways foreign partnerships shape outcomes.
DANGERS OF WEAK LOCALISATION
The consequences of weak localisation are serious. AI systems trained abroad often misidentify African faces, misinterpret African languages, and replicate systemic biases, raising concerns about discrimination and digital rights. Yet, while African AI strategies often mention ethics and human rights, we lack the institutions and consultation processes such as the six-month public consultations typical in the EU that make such commitments enforceable. As Shilongo notes, perhaps Africa should copy less of the content of Western frameworks and more of the participatory processes that make them legitimate.
In short, Africa’s AI policy moment reflects both progress and peril: policies are emerging, but without deeper local ownership, institutional capacity and participatory design, we risk entrenching dependency rather than building sovereignty.
DISPROPORTIONATE INFLUENCE OF FOREIGN TECHNOLOGY ON AFRICAN INNOVATION ECOSYSTEMS – REAL LIFE EXAMPLES
The critique of foreign dominance in Africa’s digital space is best illustrated through concrete examples that reveal how global technology companies shape local innovation ecosystems, often in ways that mirror older colonial patterns of extraction and dependency.
Language exclusion: Africa is home to over 2,000 languages (https://alp.fas.harvard.edu/introduction-african-languages > Accessed on 16th September, 2025), around one-third of the world’s total, yet, as of May 2024, Apple’s Siri, Google Assistant and Amazon’s Alexa collectively support none of them. This linguistic exclusion reinforces dependency on foreign platforms while marginalizing African cultures in the digital sphere.
Exploited labour: In 2019, South African graduate Daniel Motaung began work as a content moderator for Sama, a subcontractor for Facebook. Relocated to Kenya, he earned $2.20 per hour to review traumatic content described by colleagues as “mental torture”. When Motaung and others attempted to unionize, he was dismissed and later sued Sama and Facebook for union-busting and exploitation. This case underscores how “responsible outsourcing” in Africa often conceals exploitative labor practices.
Other News
Resource extraction: The Democratic Republic of Congo holds nearly half of the world’s known cobalt reserves, vital for powering smartphones and electric cars. In Kolwesi alone, thousands of children reportedly mine cobalt under dangerous conditions, while profits flow largely abroad. Much like colonial resource extraction, Africa provides the raw materials that power global digital economies but sees little local benefit.
Surveillance and bias: In Johannesburg, Vumacam has deployed more than 5,000 CCTV cameras integrated with AI analytics for private security firms. Activists warn that this reliance on facial recognition, already proven to misidentify darker-skinned faces at disproportionately high rates entrenches South Africa’s long history of racialized surveillance. Foreign-designed technologies thus risk reinforcing systemic inequalities under the guise of safety.
Connectivity myths: Mark Zuckerberg’s Internet.org initiative (launched in 2013) was marketed as a philanthropic effort to connect the unconnected. Projects like Free Basics promised free access to online services in over 60 countries. Yet leaked documents revealed that millions of Global South users were secretly charged for “free” data, generating nearly $100 million in 2021 alone. Framed as altruism, these projects extended Facebook’s market reach while extracting revenue from vulnerable populations.
Taken together, these examples reveal how global technology firms, mostly U.S.-based, operate in Africa with strategies that echo colonial logics. They build critical infrastructures (clouds, platforms, connectivity) aligned with their own commercial interests, entrench market monopolies and rely on low-wage labour or raw resource extraction with little local reinvestment. Their technologies often embed cultural and racial biases reflective of narrow developer demographics, yet are exported globally under the banner of “progress,” “development,” or “connecting people.”
As Western jurisdictions strengthen data protection and AI regulation, African countries often remain vulnerable due to weaker frameworks and limited enforcement capacity. This asymmetry creates fertile ground for digital colonialism; a modern-day “Scramble for Africa” where foreign firms extract and control data much like colonial powers once extracted minerals (Danielle Coleman, ‘Digital Colonialism: The 21st Century Scramble for Africa Through Extraction and Control of User Data and the Limitations of Data Protection Laws’ (Law.Umich.Edu) < https://repository.law.umich.edu/mjrl/vol24/iss2/6/ > Accessed on 16th September, 2025). Under the guise of innovation, these companies wield disproportionate influence over African AI and digital ecosystems, shaping policy choices, technical architectures, and even societal norms, while leaving Africa in a position of dependency rather than empowerment.
THE ISSUES GENERATED BY AI POLICY
While global AI policy is advancing through risk-based regulation, ethical standards, and participatory governance, Africa’s AI landscape remains fragmented, heavily modeled on external frameworks, and vulnerable to digital dependency. The disproportionate power of foreign technology companies manifested in many ways, including linguistic exclusion, exploitative labour, resource extraction, biased surveillance and deceptive connectivity projects echoes colonial logics of extraction and control. Without decisive intervention, the continent risks entrenching digital colonialism, a new form of dependency in which policy choices, infrastructures and innovation ecosystems are shaped externally, undermining both democratic values and long-term development.
WHAT AFRICAN STATES MUST DO
To avoid replicating historical asymmetries in digital form, African states must assert sovereignty over their AI policies, data governance and digital infrastructures. This requires moving beyond passive adoption toward active regulatory design, investment in local infrastructure (such as data centers, compute resources and research capacity) and strengthening institutional oversight with technically competent regulators. Equally critical is the creation of participatory policy processes that center human rights, economic development, and indigenous innovation. Only by combining legal safeguards, domestic capacity, and strategic partnerships built on equality, not dependence, can Africa transform digital technologies into engines of genuine development rather than renewed extraction.
THE NIGERIAN EXAMPLE: DATA SOVEREIGNTY OR DATA SURRENDER
With the rapid expansion of national digital infrastructure across Nigeria, a far more pressing issue has risen to the fore: the question of who truly owns and governs the data that powers this infrastructure. As digital systems increasingly underpin the delivery of public services, financial transactions, education platforms, health records, and national security functions, data becomes not only a technical asset but a core element of state power. Data sovereignty means that data generated within a country’s borders is governed by that nation’s laws and regulatory frameworks; this ensures local control over data access, storage, and usage (Folashadé Soulé, ‘Digital Sovereignty in Africa: Moving beyond Local Data Ownership’ CIGI (2024) <https://www.cigionline.org/publications/digital-sovereignty-in-africa-moving-beyond-local-data-ownership/> Accessed on the 14th of June, 2025.). It has become a critical aspect of national policy and governance. In Nigeria, this issue has grown increasingly complex, particularly in light of the pervasive presence of foreign cloud providers, offshore data processors, and international technology firms that collect, process, and sometimes export Nigerian user data without clear or enforceable jurisdictional frameworks.
Foreign digital platforms have historically played a central role in the Nigerian data ecosystem either as providers of essential services like email, storage, and analytics, or as developers of social media and financial applications used daily by millions of Nigerians (Fola Odufuwa et al., ‘Digital Technology Adoption by Microenterprises: Nigeria Report’ (2024) <https://www.researchgate.net/publication/383202125_Digital_Technology_Adoption_by_Microenterprises_Nigeria_Report> Accessed on the 14th of June, 2025.). While these platforms often promise global connectivity and technical sophistication, they also introduce serious risks. Data generated within Nigeria is frequently routed through foreign servers, stored in jurisdictions with significantly different privacy protections, and subjected to external political and commercial interests (Patrick Aloamaka, ‘DATA PROTECTION AND PRIVACY CHALLENGES IN NIGERIA: LESSONS FROM OTHER JURISDICTIONS’ UCC Law Journal (2023) 3 (1).). This dislocation of Nigerian data is what scholars term extraterritorial data flow which raises serious questions about control, privacy, and national security. The potential misuse of this data, whether for commercial exploitation, surveillance, or even geopolitical leverage, makes the issue of domestic data governance all the more urgent.
(To be continued).
Thought for the week
“Over time, I think we will probably see a closer merger of biological intelligence and digital intelligence.” – Elon Musk
