Algorithmic Populism and the Politics of Waste: How AI Reproduces Plastic Colonialism in the Global South

In this incisive analysis, Dr. Oludele Solaja interrogates how AI-driven waste governance reproduces global inequalities under the guise of efficiency. Introducing the concept of “algorithmic populism,” the article reveals how technocratic systems, framed as serving the public good, instead concentrate power within elite infrastructures while marginalizing affected communities. Through empirical insights on global plastic flows and case evidence from Nigeria, the article demonstrates how optimization logics perpetuate “plastic colonialism.” It calls for transparency, participatory design, and updated regulatory frameworks to prevent algorithmic governance from entrenching environmental injustice.

By Dr. Oludele Solaja

Even though the world was debating about a new global plastic treaty and big multinational companies were developing intelligent AI systems for managing worldwide recycling, nothing actually changed the status quo. The Global South remained the global repository for the world’s plastic waste. Far from being an outcome of ignorance or incompetence, the logic behind this persistent pattern of global environmental injustice could be explained by concepts of algorithmic populism. Algorithms designed to optimize global waste flows were simultaneously creating new forms of global environmental governance that duplicated existing power hierarchies, while ostensibly addressing a global waste crisis (Dauvergne, 2018; Brooks et al., 2018; Vinuesa et al., 2020). Algorithmic optimization, not the solution to our waste crisis, increasingly served as the vehicle for reproduction of the system of plastic colonialism in digitally encoded form.

This problem is conceptualized here by the idea of algorithmic populism. Following Mudde’s influential definition of populism as a moralized political logic that differentiates between "the pure people" and "the corrupt elite" (Mudde, 2004; Mudde & Kaltwasser, 2017), algorithmic populism suggests the new logic of governance through which algorithmic systems are promoted as apolitical tools of expertise serving the ‘people,’ yet control and authority are increasingly concentrated within a small technocratic elite (Beer, 2017; Pasquale, 2015). Within this regime of technocratic management, ‘the people‘ have been transformed into data points managed through complex computational infrastructure created and controlled by corporate and institutional entities. This structure of governance presents a facade of democratic and technical efficiency while obscuring significant inequalities in the application of decision-making authority.

This pattern reflects a wider contemporary mode of governance. As Michel Foucault noted (1980), modern power structures are built through the creation of regimes of knowledge through which what can be known and what constitutes rational and efficient behavior are determined. Within the sphere of waste governance, algorithmic systems increasingly produce their own authoritative ‘truths’ about the destinations, treatment processes and the comparative economic efficiencies of exporting or receiving waste. These truths, however, are socially embedded, shaped by a global economy in which cost efficiency may easily override concerns about environmental justice (Kitchin, 2017; Pasquale, 2015). Optimization therefore perpetuates, rather than ameliorates, patterns of global inequality.

An example of this dynamic can be observed in patterns of the global plastic waste trade. Despite international regulations such as the Basel Convention high-income countries continued to export large amounts of plastic waste into countries with limited environmental regulations (Jambeck et al., 2015; Geyer et al., 2017). When China banned imports of plastic waste in 2018, global waste flows rerouted themselves to Southeast Asia and parts of Africa, now managed through an array of global optimization, tracking and tracing algorithms that help to streamline and automate logistical operations (Brooks et al., 2018). Optimization algorithms identifying cheap destinations also naturally target locations with weaker regulatory institutions and environmental controls, typically those in the Global South.

The waste trade in Nigeria provides a clear example of this pattern. Nigeria is one of Africa’s most populous nations and one of the continent’s largest consumer markets; the nation has long faced an overwhelming plastic waste problem and is a destination country for enormous quantities of plastic waste generated both within its own borders and abroad (Dauvergne, 2018). The overwhelming majority of the informal waste picking sector in Lagos operates as an unofficial but fundamental component of waste management systems, where pickers sift through landfills and waterways for materials to recycle under dangerous and precariously employed conditions, and these workers remain completely outside decision-making circles regarding new forms of smart and algorithmic waste management (Beer, 2017; Heeks, 2022). Tools and applications developed in distant corporate and institutional settings serve to create a system of waste management that fails to account for the conditions that workers face at local sites of accumulation.

This exclusion is a manifestation of the contradictions inherent in algorithmic populism. In fact, where algorithmic governance is supposed to create more democratic forms of participation, it often works to obscure power asymmetries and lack of participation; indeed, many contemporary populist movements draw power from precisely the perception of exclusion and lack of voice, a problem increasingly amplified in the digital space (Norris & Inglehart, 2019). Environmental policy, for instance, increasingly relies on information systems and models that make decision-making opaque to even its most implicated stakeholders (Pasquale, 2015; Kitchin, 2017). As such, efficient algorithmic logic may ultimately consolidate rather than alleviate environmental injustices.

The popular circular economy model is itself a perfect illustration of this contradiction; it seeks to build a system of material flows that aims to minimize waste but ends up facilitating global waste flows through optimized systems that reproduce traditional economic and political hierarchies. As has been shown above, this circular logic simply becomes a circular illusion whereby waste continues to circulate globally in the context of unequal power relations, ultimately continuing to accumulate in the countries with weaker environmental and political infrastructure (Vinuesa et al., 2020; Dauvergne, 2018).

This difference is striking when comparing how these technologies are often experienced in different parts of the world. In Europe, AI applications in waste management are presented as "green" technological innovations, part of broader goals for climate-compatible resource consumption; in many parts of Africa, they function to exacerbate waste problems, through the continued accumulation of waste in landfills and waterscapes and increased precarious work in the informal sector (Brooks et al., 2018). Cost efficiency trumped local realities and environmental justice outcomes in Europe, while for Africa continued accumulation resulted in increased environmental degradation and precarity.

This isn’t just about failing to adequately represent the people; algorithmic populism actively digitizes populism itself. What could and should be debated as political issues around the global distribution of waste, through the processes of debate and consensus-building, are reframed and regulated as technical problems solvable through expert-driven algorithmic intervention, de-politicizing them in the process, and ushering in new forms of technocratic rule (Beer, 2017; Pasquale, 2015). Without checks on their operation, optimization-driven technologies risk legitimating environmental inequality.

There are number of solutions required to solve this problem. First, algorithmic transparency should be a central pillar of future governance of waste. Public access should be required to the decision-making logic behind algorithmic choices, including the factors used to identify destinations for waste streams (Kitchin, 2017; Vinuesa et al., 2020). Second, participatory models should be part of future design and deployment of technology systems. Waste pickers in Nigeria, for example, possess unique on-the-ground knowledge of the complex political and environmental ecology of waste that can help to create truly ‘smart’ systems that are ‘fairly smart’ and beneficial to local contexts (Beer, 2017; Heeks, 2022). Third, international governance frameworks need to adapt to address the reality of algorithmic infrastructure as a central force in shaping the contemporary global waste trade. 

Existing conventions that regulate waste flows were written prior to the rise of algorithmic systems, and new regulations and standards must be devised in order to guarantee fairness, accountability and environmental justice in technological governance (Pasquale, 2015; Vinuesa et al., 2020). Lastly, environmental technology governance needs to be de-politicized: algorithmic tools must be reconceptualized not as ‘solutions,’ but as socio-technical systems implicated in patterns of power and exclusion (Foucault, 1980). In the absence of such measures, algorithmic governance may become the ultimate tool for disguising environmental inequality as technological progress.

In conclusion, algorithmic populism reveals how ostensibly neutral technologies can entrench, rather than resolve, global inequalities. By depoliticizing waste governance and privileging efficiency over justice, AI systems risk reproducing plastic colonialism in digital form. Meaningful reform therefore requires transparency, participatory inclusion, and updated global regulatory frameworks. Without such interventions, algorithmic governance will continue to legitimize unequal environmental burdens while masking them as technical necessity and progress.

References

Beer, D. (2017). “The social power of algorithms.” Information, Communication & Society, 20(1), 1–13.

Brooks, A. L.; Wang, S. & Jambeck, J. R. (2018). “The Chinese import ban and its impact on global plastic waste trade.” Science Advances, 4(6), eaat0131.

Dauvergne, P. (2018). “Why is the global governance of plastic failing the oceans?” Global Environmental Change, 51, 22–31.

Foucault, M. (1980). Power/knowledge: Selected interviews and other writings, 1972–1977. Pantheon Books.

Geyer, R.; Jambeck, J. R. & Law, K. L. (2017). “Production, use, and fate of all plastics ever made.” Science Advances, 3(7), e1700782.

Heeks, R. (2022). “Artificial intelligence for sustainable development: The new frontier.” Development Informatics Working Paper Series, University of Manchester.

Jambeck, J. R.; Geyer, R.; Wilcox, C.; Siegler, T. R.; Perryman, M.; Andrady, A.; Narayan, R. & Law, K. L. (2015). “Plastic waste inputs from land into the ocean.” Science, 347(6223), 768–771.

Kitchin, R. (2017). “Thinking critically about and researching algorithms.” Information, Communication & Society, 20(1), 14–29.

Mudde, C. (2004). “The populist zeitgeist.” Government and Opposition, 39(4), 541–563.

Mudde, C., & Kaltwasser, C. R. (2017). Populism: A very short introduction. Oxford University Press.

Norris, P., & Inglehart, R. (2019). Cultural backlash: Trump, Brexit, and authoritarian populism. Cambridge University Press.

Pasquale, F. (2015). The black box society: The secret algorithms that control money and information. Harvard University Press.

Vinuesa, R.; Azizpour, H.; Leite, I.; Balaam, M.; Dignum, V.; Domisch, S. & Fuso Nerini, F. (2020). “The role of artificial intelligence in achieving the Sustainable Development Goals.” Nature Communications, 11, 233.