The interplay arising between Science and Technology Studies (STS), and Responsible Research and Innovation (RRI) frameworks are under scrutiny by research institutions concerned with fostering sustainable development of social technology models in the Global South. Such thematic entanglement suggests the need for a decolonial conceptual framework for technologies, asserting that Northern digital industries and corporations often operate within a colonial paradigm, thereby subjecting Southern and non-Western social assemblages (Deleuze & Guattari, 1987) to unjust, inequitable, and unreliable digital experiences and practices (Milan, Beraldo, 2024). Thus, post-coloniality (Mignolo, 2007; Escobar, 2017, Cruz, 2021; Bonami, 2022) reveals pathways to Southern emancipation from contemporary Northern and Western power structures that still dictate interactions within innovation ecosystems. Consequently, research in this intersection demands attention, particularly given (i) the rapid pace of the digital industry, (ii) the resulting imbalance among continents, (iii) the impact on marginalized and vulnerable populations exposed to unfair digital practices, and (iv) the inquiry into the feasibility of decolonizing digital technology admitting a positioned knowledge (Haraway, 1998). This paper explores the possibility of creating a methodology capable of promoting and fostering digital technology decolonization in the development of Artificial Intelligence (AI) (Eke, 2023; Russel, Norvig, 2010) systems in the Global South, drawing insights from case studies conducted in Brazil, Senegal, and Rwanda. The study will adopt a mixed-method approach to understanding a decolonial strategy in developing and appropriating AI systems, with a focus on (i) determining the contextual annotation required for a decolonial method within a given algorithm, (ii) identifying the stage at which this contextual data should be integrated to achieve diverse outcomes, and (iii) examining specific AI models that can facilitate a decolonial AI. Hopefully, these discussions will contribute to conclusions, while assessing whether digital systems can qualitatively be context-driven (Bakhouya et al., 2017) and value-centered rather than only data-driven.