Pedro Salguero 
University of Valladolid

This abstract presents the results of a qualitative study based on 12 in-depth interviews with CEOs and decision-makers in Spanish companies that outsource data processing tasks related to artificial intelligence (AI) to Latin American countries. This research is part of a larger project that examines the data value chain between Europe and Latin America, focusing on how these outsourcing practices contribute to the perpetuation of labor and economic inequalities on a global scale.

The study looks at the decisions made by companies in Spain to delegate tasks such as data cleaning, labeling, and classification—critical for AI training. Adopting a studying up approach, the study focuses on the business elites who make these decisions, rather than examining the conditions of the workers in Latin America. A key component of the methodology is the use of discourse analysis, through which the study examines how business leaders justify outsourcing decisions. 

The findings show that companies justify outsourcing primarily in terms of efficiency and cost reduction, framing these practices as standard business strategies that enhance global competitiveness. However, behind the corporate rhetoric about process optimization and technological progress, the social and labor implications are obscured. These practices continue to exploit Latin American workers, highlighting the contrast between technological advancement in the Global North and the exploitation of workers in the Global South.

The study also reveals how Spanish companies construct narratives around outsourcing that present it as a neutral, rational, and cost-effective strategy, while overlooking the colonial structures inherent in these decisions. These narratives are further laden with stereotypes about Latin American workers, who are seen as less efficient and in need of constant supervision. Meanwhile, both companies and governments seem far more focused on cybersecurity and data protection—benefiting private and public interests alike—than on the labor conditions in the places where these data are processed. This disconnection between corporate policies and local realities reveals a systemic indifference to the situation of workers in Latin America.

Through a critical analysis of corporate narratives and the underlying power structures, this study demonstrates how outsourcing AI services between Europe and Latin America not only perpetuates an economic exploitation model but also strengthens the persistence of colonial relationships in the global digital economy. Far from representing a move toward global equity, these practices emphasize the geopolitical and labor inequalities that continue to shape work dynamics in the 21st century.