Challenging the Coloniality in Global AI Regulation Frameworks

Abstract

This blog argues that the intensifying efforts to regulate AI, notably through initiatives like the EU AI Act, are predominantly driven by the Global North, often neglecting the significant socio-cultural, legal, and economic differences across countries. While there is widespread consensus that AI governance must adopt a global approach due to the technology’s borderless nature, many Global South nations are uncritically adopting AI regulations modelled on those from the Global North. This tendency risks reinforcing colonial mindsets, as these frameworks often fail to account for the diverse ways AI impacts different societies. The blog calls for a more nuanced and context-sensitive approach to AI regulation that challenges these existing power dynamics and better reflects the realities of post-colonial nations.

Keywords – Decolonialism, AI, Governance, Regulation, Lawmaking, Global South

Introduction

In present times, regulation of Artificial Intelligence has become a very critical issue. This is also at a time when nations in the Global South as well as in the European Union are taking steps to formulate rules for governing the same. AI regulation is fast becoming one of the most critical global issues, with nations in the Global North (GN), like those in the EU, taking the lead in shaping the rules of the game. But here’s the problem: these efforts, such as the EU AI Act, are paving the way for international standards. They often fail to account for the vastly different realities of countries in the Global South (GS).We argue that post-colonial nations in the GS too often adopt these frameworks without adapting them to their own socio-cultural, legal, and economic contexts.

The heart of the issue goes beyond modern-day power imbalances.   This kind of power imbalance does not lead to simply enforcing international standards. It also reflects a deeper and troubling history.  During the colonial era, this kind of a system used to exist, where the governance system from developed nations were imposed on native countries, without taking into consideration their unique contexts. In present times, the adoption of AI laws by countries in GS perpetuates the idea of continuing colonial legacies. Such kind of regulatory models adopted from GN are not reflective of the existing needs of native societies.  AI also presents different kinds of challenges for diverse societies.  AI systems are usually trained on those datasets which reflect the existing norms of GN, which are usually not suitable for the countries which are situated in GS. This can lead to various issues, from cultural misrepresentation and appropriation to reinforcing stereotypes. Furthermore, ethical considerations around AI differ from region to region. In some areas, deploying AI technologies may clash with local beliefs, traditions, and social structures, affecting how AI is received, regulated, and governed.

While legal frameworks are essential in addressing broad issues such as bias, privacy, fairness, and accountability, they must be context-sensitive. A one-size-fits-all approach to AI governance is unlikely to succeed, particularly when the socio-cultural realities of post-colonial societies are not adequately considered. We advocate for a governance structure that is not only legally robust but also culturally sensitive—one that respects the values and aspirations of the communities it seeks to regulate. AI laws must be adapted to local contexts to avoid perpetuating the colonial mindset that continues to shape global power dynamics.

Decolonial Perspectives on Power, Data, and Governance

Decolonial perspectives on AI are not just abstract intellectual exercises—they are a bold challenge to the continuing colonisation of the GS by the very AI systems meant to drive progress. These perspectives force us to confront the uncomfortable reality that AI is not neutral; it is steeped in the biases, power dynamics, and inequalities of its creators.At its core, decoloniality serves as a radical lens that exposes and deconstructs the deep-rooted colonial structures still shaping our world. It calls for reclaiming and amplifying knowledge systems silenced by colonialism, challenging the dominance of Eurocentric narratives.

The idea of ‘coloniality of power’ makes one thing clear: colonialism is not a relic of the past—it is a living, breathing force that still shapes our economies, cultures, and social systems. It does not just conquer land; it colonises minds, erasing alternative ways of thinking and being. Far from being over, colonialism has evolved, embedding itself in the very structures that uphold the global inequalities between the GN and the GS. It is still here, reinforcing racial hierarchies and maintaining the dominance of Western knowledge, quietly steering the world in its favour.

AI, through this lens, becomes a tool for maintaining these colonial legacies. Far from being a benign or purely innovative technology, AI is another instrument of domination. Developed primarily in the GN, AI systems carry the biases, assumptions, and power structures of their creators. However, in our view, this goes beyond a simple technical issue; it reflects a continuation of colonial dynamics, with the GN once again asserting its frameworks and systems over the GS. The extraction of data mirrors the historical exploitation of land and resources. Just as colonisers dismissed Indigenous peoples and claimed their territories, today’s corporations treat data as a new frontier to be seized and monetised, often ignoring its human and societal impact. Hence, it becomes evident that data colonialism is the digital continuation of colonial domination. BigTechs exploit data in ways eerily reminiscent of the resource grabs of the colonial era. This “data grab” turns the global digital economy into yet another vehicle for reinforcing power imbalances, where the GN once again claims to act for the “greater good.” But, as always, the benefits flow disproportionately to the GN, deepening the economic and intellectual divide with the GS.

The myth of ‘data-centric rationality’—that data offers a neutral, objective reflection of reality—crumbles under scrutiny. The entire architecture of data collection, analysis, and control is dominated by GN institutions operating within a capitalist framework that prioritises profit and power. We argue that it represents a reconfiguration of power dynamics tailored to the digital age. The “datafication” of society, especially in the GS, flattens rich local cultures and erases complexities, all to serve the interests of a powerful elite. However, the colonial reach of AI goes beyond mere data extraction. AI systems today operate as tools of modern colonisation, imposing external norms and standards on local communities while ignoring their cultural and social realities. This goes beyond flawed algorithms—it is about sustaining the very structural inequalities rooted in colonialism. Technologies like facial recognition, which consistently fail to recognise darker-skinned individuals, or predictive policing systems that disproportionately target marginalised groups are not accidents. They are symptoms of a broader system designed to uphold and reinforce colonial hierarchies, no matter how advanced or “innovative” it pretends to be.

Even AI ethics guidelines, particularly those from GN corporations, are steeped in these power dynamics. While initiatives like UNESCO’s Recommendations on the Ethics of AI offer some hope, many ethical frameworks are designed by the very corporations that benefit from the status quo. These frameworks often serve as tools for “ethics whitewashing,” using the appearance of ethical compliance to conceal deeper systemic issues. What’s urgently needed is a decolonial approach to AI ethics and governance. Scholars like Gwagwa et al. and Mhlambi  argue that AI must incorporate Indigenous knowledge systems, such as the African philosophy of ‘ubuntu.’ Ubuntu, which emphasises interconnectedness and collective care, offers a framework for building AI systems that are more just and equitable, addressing historical wrongs and including marginalised voices in the conversation.

Decolonial theories face their critics, who claim they oversimplify the complexities of Western societies and cast Indigenous cultures as static. But the uncomfortable truth is that colonial legacies, like the doctrine of discovery, still shape modern governance—especially in areas like property rights and resource extraction. These same colonial patterns are playing out in the digital world, where data is treated as a resource to be exploited, just waiting for GN corporations to claim it. While there is an argument that datafication lacks the apparent violence of traditional colonialism this misses the point. Modern dispossession is more subtle but no less harmful. Just as colonialism robbed Indigenous peoples of their land and resources, today’s data economy extracts value from marginalised communities without their consent, reinforcing entrenched inequalities. The idea that data has no value until it is “discovered” and monetised by GN corporations is a digital echo of the colonial mindset that saw GS resources as worthless until outsiders took control.

Challenging Colonial Power Dynamics in Legislative Framework

The traditional legislative models, often celebrated as neutral, fair, and reasoned, remain deeply influenced by colonial legacies, which distort both the motivations behind and the effectiveness of AI regulation. The ‘rationalist’ approach, grounded in logic and evidence-based decision-making, presents an idealised image of laws designed for the common good. Yet, this model is far from impartial. As discussed before, historically, the law served as an instrument of imperial dominance, used by colonial powers to impose foreign legal frameworks on subjugated nations under the guise of a “civilising mission.” Today, post-colonial nations are still entangled in these inherited legal structures, navigating a complex web of cultural and social legacies left by colonial rule.

In the context of AI governance, the shortcomings of the rationalist model are particularly stark. Its reliance on expert-driven processes disproportionately favours the GN while disregarding the local knowledge and lived experiences of the GS. This results in an over-simplification of complex, context-specific issues into quantifiable metrics, missing the cultural, historical, and social nuances that shape the impact of these laws. The GN’s legal frameworks, often presented as universal standards, risk being imposed on the GS in what can be described as a modern form of ‘governance colonialism.’

This dynamic mirrors the colonial doctrine of ‘terra nullius,’ where indigenous lands were deemed “empty” and thus available for exploitation. Today, the GN’s legislative models for AI are similarly promoted as universally applicable, with little regard for the unique contexts of non-Western nations. This reinforces the coloniality of power, positioning European legal systems and knowledge as superior, while non-European frameworks are dismissed as incapable of producing “rational” or legitimate alternatives. The result is a continued marginalisation of non-Western perspectives and a stifling of cross-cultural legal exchange.

The illusion that European frameworks represent a universal standard for progress continues to captivate post-colonial nations, perpetuating a form of ‘internal colonialism.’ These countries replicate foreign legal systems that are ill-suited to their needs, thus reinforcing the very colonial structures they should be dismantling. This dynamic upholds the Western legal hegemony at the expense of locally informed legal approaches that could more effectively address the unique challenges of AI governance in the GS.

However, a more critical framework—the conflict model—reveals the inherent power dynamics at play in the legislative process. Unlike the rationalist approach, which assumes laws emerge from a place of neutrality, the conflict model highlights how legislation reflects the interests of dominant groups. In AI governance, this means that laws tend to favour the nations, corporations, and entities that wield the most power—often the GN—while the GS struggles to assert its interests. The conflict model unveils the persistent inequalities embedded in international AI regulatory frameworks, wherein the GN continues to shape global governance, often at the expense of the GS.

When examined through a decolonial lens, the conflict model not only underscores the power imbalances inherent in law-making but also shows how these imbalances are deeply tied to the historical legacies of colonisation. In this sense, AI governance becomes an extension of colonial power structures, wherein the GN maintains its dominance through control over legal standards, technological advancements, and access to resources. The result is a widening gap between the GN and GS regarding technological influence and legislative agency.

Thus, the central question for post-colonial nations is not whether to adopt the GN’s legal frameworks but whether to challenge these structures and advocate for laws that reflect their distinct social, cultural, and political realities. The assumption that AI laws from the GN no matter how well-researched, can be uniformly applied to the GS is not only misguided but deeply colonial in nature. Post-colonial nations must assert their own legal sovereignty, moving beyond the passive acceptance of external “gold standards” and towards creating contextually relevant and inclusive AI governance frameworks.

Conclusion

 It has become extremely important that regulation of AI brings out its true potential, which maximises the benefits for society. The regulation must also minimise the impact of its negative effects on society. This piece has argued that the drafting of regulation of AI cannot be done solely by a handful of regulators in the Global North. This is due to the fact that they often fail to take into account the unique scenario of countries situated in the Global South. Any AI governance must be done through a “radically decolonial lens,” which must be responsive to the needs and realities of the Global South.

A decolonial approach towards the regulation of AI demands that it questions existing legal colonial legacies which are deeply imbibed in AI and data governance. It offers a new path which is bereft of any kind of existing inequalities present int the digital era. It offers a transformative path towards dismantling the entrenched inequalities of the digital era. Ultimately, AI regulation must serve as a catalyst for a future, which brings prosperity. It must not consolidate the dominance of a few powerful people.  

In order to truly benefit from AI, the Global South must tread its own path, which recognises its own unique history, sociological background of the people and surrounding politics. In such a way, there must be a promotion of legal autonomy by Global South. A more inclusive and just global order is a need of the hour. The future of AI lies in addressing the imbalances which have been perpetuated for a long period of time.

This article is a part of the DNLU-SLJ (Guest Post) series, for submissions click here.

Professor Subhajit Basu & Adekemi Omotuboraa

Professor Subhajit Basu is a Professor of Law at Leeds University. He was the Chair of the British and Irish Law Education Technology Association (2018-2021). In 2020 he was awarded the Hind Rattan (a Hindi phrase which translates to English as “Jewel of India”) by the Non-Resident Indians (NRI) Welfare Society of India, Government of India, for his outstanding contribution to education and achievements in the field of Information Technology Law.

Adekemi Omotuboraa holds phd and master degrees on information technology law from the University of Leeds, UK and currently teach law at the University of Lagos, Nigeria. She specialises in investigating the law and regulation of emerging technologies particularly in developing countries.