Decolonial AI: Decolonial Theory as Sociotechnical Foresight in Artificial Intelligence

By Shakir Mohamed, Marie-Therese Png, William S. Isaac. 2020

669 citations recorded by the source corpus.

This paper explores the important role of critical science, and in particular of post-colonial and decolonial theories, in understanding and shaping the ongoing advances in artificial intelligence. Artificial intelligence (AI) is viewed as amongst the technological advances that will reshape modern societies and their relations. While the design and deployment of systems that continually adapt holds the promise of far-reaching positive change, they simultaneously pose significant risks, especially to already vulnerable peoples. Values and power are central to this discussion. Decolonial theories use historical hindsight to explain patterns of power that shape our intellectual, political, economic, and social world. By embedding a decolonial critical approach within its technical practice, AI communities can develop foresight and tactics that can better align research and technology development with established ethical principles, centring vulnerable peoples who continue to bear the brunt of negative impacts of innovation and scientific progress. We highlight problematic applications that are instances of coloniality, and using a decolonial lens, submit three tactics that can form a decolonial field of artificial intelligence: creating a critical technical practice of AI, seeking reverse tutelage and reverse pedagogies, and the renewal of affective and political communities. The years ahead will usher in a wave of new scientific breakthroughs and technologies driven by AI research, making it incumbent upon AI communities to strengthen the social contract through ethical foresight and the multiplicity of intellectual perspectives available to us, ultimately supporting future technologies that enable greater well-being, with the goal of beneficence and justice for all.

Subfields: Cognitive Neuroscience, Information Systems, Safety Research, Safety, Robustness & Interpretability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Information Systems

In Safety Research

In Safety, Robustness & Interpretability

Citations in the corpus

Ethical and social risks of harm from Language Models (2021, background): “Beyond publishing research, we believe responsible innovation also requires inclusive dialogue between stakeholders in AI development which includes affected communities and the wider public (Gabriel, 2020b; Mohamed et al., 2020; Murgia, 2021; Stilgoe et al., 2013).”
Re-imagining Algorithmic Fairness in India and Beyond (2021, background): “, in non-consensual human trials and waste dumping [123] (also see [137]).”
Evaluating the Social Impact of Generative AI Systems in Systems and Society (2023, background): “Global deployment of a model can consolidate power within a single, originating culture, to determine and propagate acceptability [245] across cultures [150].”
Process for Adapting Language Models to Society (PALMS) with Values-Targeted Datasets (2021, background): “[24] analyze the power hierarchies among groups developing AI policies in a global context, demonstrating the need to include marginalized voices in the policy development process.”
Predictability and Surprise in Large Generative Models (2022, methodology): “Along with innovations in governance of AI organizations, and work on best practices, we also believe governments should invest in better methods to assure the benefits of systems being deployed - specifically, governments should support efforts to measure and monitor the capabilities (both harmful ”
Data and its (dis)contents: A survey of dataset development and use in machine learning research (2020, background): “Particularly in consideration of an international context, where privacy concernsmay be less regulated in certain regions, the potential for data exploitation is a real threat to the safety and well-being of data subjects.(104) As a result, those that are the most vulnerable are at risk of losing co”
Data Governance in the Age of Large-Scale Data-Driven Language Technology (2022, background): “[95] Shakir Mohamed, Marie-Thérèse Png, and William S.”
Characteristics of Harmful Text: Towards Rigorous Benchmarking of Language Models (2022, background): “Finally, every application is shaped by a social context [60, 77, 92, 53], which includes a range of factors such as language and cultural norms in the community using a system.”
Power to the People? Opportunities and Challenges for Participatory AI (2022, background): “the net benefit of participants, especially those disproportionately impacted by oppressive social structures, who often carry the burdens of negative impacts of technology [12, 22, 62, 65] and reflecting a fair or proportionate return for the value of the participatory engagement.”
The uselessness of AI ethics (2022, background): “One specific strain of work has begun to think more concretely about ways to decolonise AI, unraveling histories of inequality and asymmetric systems of power [46].”
Expanding Explainability: Towards Social Transparency in AI systems (2021, background): “Otherwise one may risk deploying an AI system un-integrated into individual and organizational workflows [66, 106], potentially resulting in misuse, mistrust [108, 109], or having profound ethical risks and unintended consequences, especially for marginalized groups [72, 86, 99].”
Ethical Machine Learning in Healthcare (2020, background): “Finally, we note that machine learning also could and should be harnessed to create shifts in power in health care systems (148).”
Black Feminist Musings on Algorithmic Oppression (2021, background): “technical practices take a middle ground between the technical work of developing new AI algorithms and the reflexive work of criticism that uncovers hidden assumptions and alternative ways of working” [25].”
The Forgotten Margins of AI Ethics (2022, background): “Although there have been notable attempts to incorporate non-Western and feminist care ethics into the existing ethical paradigms within AI Ethics and Western philosophy [56, 76], these attempts remain marginalized by the structure of both disciplines.”
Decolonising Speech and Language Technology (2020, background): “But it also brings unhelpful assumptions about goals and methods, and a habit of algorithmic and technological colonisation (Grossman, 2014; Mohamed et al., 2020).”
Generative Art Using Neural Visual Grammars and Dual Encoders (2021, background): “We should also be conscious of some biases in the multimodal transformer [34][35].”