Introduction:
Artificial intelligence (AI) is slowly entering policing, welfare support, and the legal systems of the Global South, thereby changing how institutions operate. The introduction of AI in the Global South is not only a technological phenomenon of modernization of the equipment, but also a complete restructuring of the government, which was realized by replacing the current infrastructure. Without transparency and challenge mechanisms, the locus of accountability will move not only out of the formal oversight structures, but also into more opaque processes.
Methods:
The research will use a mixed-methods approach to compare the acceptance of AI systems in India, Brazil, and Kenya. It analyzes 35 procurement contracts, AI strategy studies on the national level, and statistical data of predictive policing and welfare automation. Those findings are critically triangulated against 24 semi-structured interviews with judges, public defenders, technologists, and representatives of civil society. In place of gauging the effects on efficiency, the analytical framework focuses on the conversion of operational authority, changes in discretionary space, and the stability of procedural protection within institutional practice.
Results:
It has been found that algorithmic infrastructures mechanize the operational power of proprietary technical ecosystems, thus restricting the discretionary judgment and institutional access to model construction and information logic. Respondents identified a few avenues to challenge automated decisions, noting inconsistencies in regulations. The comparative analysis of the three national settings shows that the tendency to refer to outside vendors and spend minimal internal audit power is universal.
Conclusion:
The introduction of AI in the Global South is not only a technological phenomenon of modernizing equipment, but also a complete restructuring of government, achieved by replacing the current infrastructure. Without transparency and challenge mechanisms, the locus of accountability will move not only out of the formal oversight structures, but also into more opaque processes.
