ABSTRACT
The growing awareness of the harms associated with developing and deploying intelligent systems triggered the emergence of ethical frameworks aimed at inhibiting these adverse effects. From a decolonial perspective, this chapter argues that ethical frameworks developed by powerful actors are problematic if they do not address power asymmetries and structural and epistemic violence. To identify the limits and implications of applying ethical frameworks that do not address the problem of power and the role Artificial Intelligence (AI) technologies play in reproducing structural systems of oppression, it proposes the dimensions that we must consider to develop AI ethics and technologies with a decolonial and feminist perspective. Also, it reviews some examples of the problem of establishing ethical frameworks to guide the governance, design, and deployment of AI systems without considering the impacts on specific contexts and the harm beyond the place they are designed. Thus, it explains why and how ethical frameworks must consider diverse epistemologies and avoid assuming that the underlying system of values in ethical frameworks developed in Western cultures is universal, expanding a decolonial imagination that explores the possible forms of AI decolonization toward the construction of technodiverse futures.
The ultimate expression of sovereignty lies, to a large extent, in the power and ability to dictate who may live and who must die.
Achille Mbembe