Artificial Intelligence and the Circular Economy in Africa: Key Considerations for a Just Transition


The decoupling of economic growth from environment degradation through more sustainable  consumption and production policies has become one of the core pillars of many post-pandemic economic recovery strategies across the globe. Consequently, the circular economy (CE) model has gained traction as an alternative to the “take, waste, and dispose system” associated with the linear economy, which for centuries has supported  imperialist economic activity across the globe and shaped economic activity across the world since the beginning of the Anthropocene

While the concept of circularity is not new in Africa, in addition to this, there is extensive literature on its application in indigenous communities. Furthermore, there is growing hype that the potential industry-wide applications of artificial intelligence (AI) for various use cases can enhance circularity in Africa. The expectation is that AI-led circularity will result in inclusive and sustainable industrial development and a more equitable, resilient, and sustainable recovery – ultimately mitigating the negative environmental outcomes associated with the linear economic model. But will it?

While technological disruption have often facilitated industrial and technological revolutions and their associated shifts in socio-economic activity, these disruptions have disproportionately benefitted many wealthier economies, often with little regard for the negative impacts they have for the most vulnerable in these ecosystems. The latest technological revolution – digital – despite having the potential to facilitate more sustainable growth trajectories, also presents systemic risks that could emerge as the result of this diffusion. This could also contribute significant negative environmental impacts to the ecosystems where they are deployed.

While the discourses on AI-based “green new deals” and circularity coupled with AI are still in early stages, leveraging AI to enhance circularity and sustainability will succeed only if it can also support the goal of a dignified life for all humans, and facilitate a just transition for all countries. To do this AI needs to be used in a manner that acknowledges the historical, cultural, political, economic, and technological contexts where AI-led CE models are implemented.    

In Africa, we need to consider the following realities:

Limited Evidence that AI can Support Inclusive Circularity in Africa

With a few exceptions, most  countries well positioned to build AI-based circular innovations that can support systemic change are based in the Global North. These countries have ecosystems where communities, governments, and private actors have the requisite endowments and enablers that facilitate technological-enabled responses to climate and ecological change. They are able to draw on their  infrastructure and resources to develop strategies that facilitate environmentally friendly productive activity, waste reduction, and efficient consumption across the entire supply chain network in key industries.

A matter of concern is that although advocacy for a circular-based economy and its implementations through AI is increasing globally, most of the research supporting this advocacy has been conducted in high-income countries. Little is known about the conditions necessary to advance the use of AI in support of the CE in African contexts, and the barriers to implementation. Some critical issues for consideration include infrastructure, technical capacity, digital capabilities, and enabling partnerships, as well as the lack of reliable data in key areas such as consumption and waste. Practical use-case examples of how an AI-led CE can be applied in economic realities that are in stark contrast to those found in wealthier nations are necessary.

AI-related Risks

While the deployment of AI-based technologies is critical for sustainability, there is limited attention to the possible systemic risks that AI presents to the advancement of just and equitable sustainable development, including:

(1) algorithmic bias;

(2) uneven access and benefits;

(3) worsening inequities and external disruptions;

(4) the environmental impacts and energy requirements of AI technologies themselves; and

(5) trade-offs between productive efficiency, economic justice, and ecological resilience.

The true impact of AI in creating a sustainable economy can only be realised if the potential systemic risks are identified, and the structural conditions needed to support more just and sustainable digital economies and futures are considered. To date, these have not been the focus of AI-governance tools. In order to achieve the Sustainable Development Goals (SDGs), regulations in various value chains, the responsible development and deployment of AI in a manner that fosters equality and inclusion, and the shared prosperity for all participants.

Uneven Transgeographical Technological, Environmental, Social and Governance Factors

While leveraging AI in a responsible manner can be a necessary condition to develop and sustain CE initiatives, the challenge of closing global materials loops and regenerating natural assets by using technology is a daunting task with many dimensions to consider. The linear or “take-make-dispose” economy that has shaped industrial evolution in many Global North countries has long relied on structures that ensure cheap and available factors of production (labour, land, capital). These were needed to create conditions for industrialisation-led economic growth and ancillary phenomena, while global capitalism-based institutions have also facilitated industrialisation and technological innovations in these countries and were central to economic growth and development for those who manage and control today’s linear supply chains and global production systems.

While there are pockets of robust,  localised production in developing regions, we cannot ignore the uneven technological, environmental, social, and governance (TESG) power dynamics such as: geo-political influence, global division of labour, technological endowments, control of value chains, and economies of scale that exist in the increasingly data-driven global economy. In addition, the complex and interconnected global risks of resource competition, commodity price volatility, and changing consumer demands, amongst others, highlight Africa’s vulnerability in these systems.      

Acknowledging global interdependencies, developed countries continue to pressure developing countries to adopt so-called “global standards, policies, and institutional arrangements” with visions for leveraging AI for sustainable utopias that are inadequate to address the everyday realities and struggles of the Global South. These visions often blatantly ignore how imperialist the policies and institutions driven by rich countries are and whether these interventions are suitable and appropriate for developing countries.

Decolonising Environmental Stewardship and Creating Just Knowledge Production

Industrialisation-based environmental destruction, systemic oppression, and the climate crisis are interrelated and are a perpetuation of colonial legacies that are inherent in the global structure of multilateral institutions and systems that shape the global economy. Also, historical legacies still the discourse on environmental stewardship, and ownership of resources is still shaped by global conservation institutions and policies that, despite several civil society initiatives, largely exclude and discriminate against indigenous and rural communities.

Decolonisation in environmental stewardship is needed to strengthen and improve existing narratives about ecological conservation by recognising that diversity, equity and inclusion in knowledge production broaden fields of knowledge by including traditionally excluded perspectives in the global governance and policy agendas. Consequently, the normative idea of “green new deals” has raised complex questions regarding who has the expertise to frame alternatives in a diverse world and the implications of these alternatives for marginalised people in different socio-economic contexts. Just knowledge production is needed to frame development paradigms that draw out the methods and conceptual frameworks that consider the realities of underrepresented actors in the discourse.  

Diverse perspectives are needed to initiate a truly equal dialogue where we explore emerging AI risks, identify critical questions, and discuss the limitations of current governance mechanisms. This is necessary to examine under what conditions an AI-based CE will be able to reduce global inequalities and promote human development in the context of the SDGs in Africa. Also, the inextricable link between gender, racial, ethnic, social, and climate injustices needs to be considered in these processes.

The takeaway is that, while leveraging AI to enhance circularity can provide possible solutions to cross-cutting development issues that Africa faces, these innovations may not always result in a positive outcome. They depend on the framework in which AI is used and the existing ecosystems where it is deployed. We need to critically assess current narratives and emerging pathways of both AI and circularity in the Global North and the Global South, and understand how these differ or possibly even conflict with one other.

The main objective of the beneficial AI and sustainable development theme of the Africa Just AI Project is to contribute towards research on AI and the SDGs through a focus on circularity in Africa. To this end, as part of supporting just AI through research, partnerships, practical application, and advocacy, the African Circular Economy Network (ACEN) and Research ICT Africa will collaborate to shape the African research agenda, identify key themes, and frame questions for further exploration to understand if and how beneficial AI can support circularity and sustainable development in Africa.