Home Essay Decolonisation and GenAI: Are We Decolonising AI or Automating Coloniality?

Essay

Decolonisation and GenAI: Are We Decolonising AI or Automating Coloniality?

April 10, 2026

750

Early in 2026, I agreed to write an essay on decolonisation and AI for the Bristol Institute for Learning and Teaching (BILT). The original article was published on the 23rd of March and can be found here. I have added many thoughts to the original text in the essay below, in which I question the possibility of “decolonising AI”. My grounds for this include the fact that decolonisation and the push to adopt GenAI in everything have diametrically opposed logics, means and aims. A caveat: I am writing specifically about the adoption of GenAI in the context within which I work.

Since the recent rise in popularity in the use of generative AI, a number of scholars have tried to conceive of ways in which it could be beneficial to decolonial motives or in which decolonisation could improve it. This essay is a reflection on the intersection between these two fields and their adopters. I am particularly concerned here with what questions are raised at that intersection and what those tell us about the nature of human existence in the modern world.

First, to clarify two broad terms used here – “AI” and “decolonisation”. In the context of this essay, I am thinking about a narrow band of “artificial intelligence” while still commenting on the proposition that computers can in any way simulate human learning, comprehension, problem-solving, decision-making, creativity and autonomy. In particular, I am interested in the ways it is suggested that computer technology can autonomously simulate and originate genuinely human approaches to teaching, research and administration. I focus on this because the use of generative AI begins to produce widespread controversy when it is argued that AI can do things (considered inherently human tasks) like write an original academic article or design bespoke teaching. In essence, what is in question is the idea that AI can autonomously replicate/originate human thinking and not just assist it. This also overlaps with the question of whether, in the academic field, things like essays and teaching are merely useful products or also inherently evidence of useful processes.

In what ways can AI interact with the decolonial movement?

The speed at which computers work has always been an attractive aspect of their use. This is important when our institutions demand more work out of us in the same amount of time, using half the amount of resources. So, people interested in bringing decolonisation into the university sector have sometimes suggested that AI’s speed could result in increased accessibility to and analysis of information and archives that are out of reach of marginalised communities. Additionally, AI could help preserve dying or extinct languages impacted by the colonial project. It could also aid in decolonising museum collections which contain artefacts stolen and/or taken during colonial occupation. Where scholars wish to adopt decolonisation and have limited expertise in this, AI has been considered an effective tool to identify gaps. These sound like admirable goals. But are they wholly/purely decolonial ones in the context of AI use? What are the points of conflict?

But what does decolonisation really mean… again?

It is often the case that discussions of decolonisation focus mainly on concepts such as equality, diversity, inclusion, and representation. These are definitely good goals and tools for our institutions, but do they amount to decolonisation if we do not disrupt the significant harm that the colonial project has enacted and, in many ways, continues to enact? (In this context, it is also important to ask where AI sits: within or outside these harms?). I suggest that it is impossible to even begin to effectively engage with the project of decolonisation if we do not understand quite clearly the structure of the project it responds to and seeks to dismantle – colonialism. Also important in this is a clear understanding of the tools of the colonial project.

Coloniality of Power as defined by Anibal Quijano

Colonialism as an ongoing ideology and structure is driven by logics of extraction, exploitation and dispossession, powered by the manufacture of artificial human hierarchies – mostly racial ones. Exploitative and extractive empires introduced in the late modern European period are those characterised by the logics of commodification of space and nature, as well as (through the manufacture of the social category of race) the commodification of humanity and variably valued labour. Built on these overarching logics is the mostly racial and gendered categorisation and hierarchisation of peoples into those who labour and those who benefit from that labour. Colonialism is thus a world-making project, and the ontology of Euro-modern law assists the enduring reproduction of this project. What I mean is that within Euro-modern law, bodies and space-time are revalued and redefined to meet the needs of enclosure, property-making and capital accumulation. The colonial world creates its own reality that directs our human existence. What it means to be human. What it means to exist on the earth. What it means to pass through time. All these are determined by the perpetual motion of the extractive colonial machine.

What then is decolonisation?

Decolonisation responds to and refuses the forceful imposition of destructive ways of living and being human in the world. It can thus be described as the long-term and continuous repudiation of colonial practices and effects, which has manifested differently in different parts of the world. In essence, decolonisation seeks a myriad of flourishing ways of living and being human in the world… ways that are not extractive, exploitative or destructive. In other words, decolonisation is not one thing but a set of context-dependent strategies adopted by those resisting colonialism – strategies specifically relevant to the particular ways in which colonial ideologies manifest themselves in those particular places. In practice, this has often involved indigenous peoples, colonised peoples and racialised peoples taking up the tools that they have to resist the specific forms of oppression that they experience at the time that they experience them. The tools of decolonisation include demanding sovereignty and reclaiming autonomy over being and resources, as well as reviving lost epistemologies. Decolonisation seeks ways of being human once more by refusing the exploitative techniques of colonial pseudo-being.

What then do we find at the intersection of AI and decolonisation?

First, why do some people hate AI so much? One reason is that GenAI promises more than it delivers. In fact, it promises a very different thing from what it actually delivers. Large Language Models (LLMs) in particular are trained on massive text datasets (often without permission). So apart from anything else, it relies on theft. LLMs aim to predict, generate, and process language. So its proponents claim that AI can “understand” context and recognise patterns; thus, it can engage in conversation, content creation, summarisation, and translation.

But it is important to be clear on what GenAI such as this is actually doing in academia. It is not thinking; it is at best simulating thinking. It is not creating; it is replicating already thought thought. It is not sentient, so it cannot be creative or imaginative. It cannot feel or think. Because of this, GenAI just outrightly lies a lot of the time. GenAI hallucinates because it aims to predict the most likely next words in a sentence rather than verify facts; it frequently values being helpful over being accurate. These models are trained on large amounts of internet data—which can be unreliable—and may sometimes use strategic deception to achieve objectives, even pretending to be aligned or safe when they’re not. This is just not how human brains work. But colonial pseudo-humanity is close to this. Lies and theft.

In other words, GenAI very often relies on stealing intellectual property and human/natural resource use and abuse that has hit those already marginalised by the colonial project the worst. It simulates already existing dominant human thinking in a colonial world, so it replicates and sometimes deepens already existing bias in deeply misogynist and racist ways. There are also many documented cases where people who use it have been encouraged by it to take their own lives.

If any other office appliance did so much harm, it would have been chucked out of the window. Yet we talk about GenAI like it is human and inevitable, as if it can blur the lines between human and machine, between human beings and office tools. But the human thinking it claims to replicate is colonial thinking – extractive, destructive, exploitative and unimaginative. We are destroying our planet, our lives, our health – for what? The ability to produce a hallucinated summary quicker and avoid the work of critical and actual human and actually generative thinking?

What is genuine “human” thinking, and can AI do it?

In our engagement (or not) with GenAI, we must remember that one of the devices of the colonial project was to make being human for most of the world an unattainable goal. Decolonisation, as an anticolonial approach, tells us that we can reclaim being human again, we can reclaim the earth, and we can reclaim time – not in extractive, exploitative and destructive ways but in flourishing ways. Thus, being genuinely human means being in community with each other and all living things. It means taking feeling seriously as anticolonial praxis – Eslava invites TWAILers to “build on feeling-thinking-doing a more just planet for all”. Christine Black argues for feeling as fundamental to a just jurisprudence for our world. Being human once more in anticolonial ways is a praxis of refusal and hopefulness, and that action develops and evolves by constant caring/feeling relations with other beings and the earth upon which we all currently precariously survive. It relies on the non-exploitative reclamation of land and time. It refuses the ways in which our being human is exploited and ignored. So that we have space to grow and time to slow down and think and create. Together. For each other. For all of us. So that we can together dream of new, flourishing and beautiful anticolonial worlds from within our institutions. And do the thinking that builds them. We need to do that thinking. We. Us. In time. With time. For each other. With each other.

Through environmental devastation, massive inequality and all forms of violence, survival is being threatened on a planetary scale. To survive, we desperately need new ways of thinking, doing and being in the world. The world desperately needs a body of knowledge that gives us a way “to see, to hurt, to feel, but also a way to dream, to move, to sit still. To live in relation and not possession. A way to be.” To be. To be human. Still. Once more. Despite everything that has gone before.

Reclaiming our collective humanity and making beautiful worlds together will be achieved not by abdicating our responsibility for doing the required new thinking and building to office equipment or a lying word-juggling machine. We are still human, for now. There is still time to think and build human worlds. While we become human once more.