Labour And Exploitation Processes in Artificial Intelligence: Example of Digital Taylorism in Data Labelling

The development of artificial intelligence (AI) heavily relies on vast datasets that require extensive human labour for annotation and labelling. Despite AI's portrayal as an autonomous and intelligent system, its functionality is deeply dependent on precarious digital labour, particularly in the Global South. Data labelling, a critical process in AI training, is often outsourced or crowdsourced, subjecting workers to low wages, job insecurity, and exploitative working conditions. This study examines the intersection of AI, labour, and global inequalities, highlighting how Digital Taylorism and datafication intensify worker surveillance and reduce human labour to fragmented, repetitive tasks. The rise of digital sweatshops and a digital underclass further illustrates how AI development perpetuates historical patterns of economic dependency between the Global North and South. Through an analysis of crowdsourcing platforms and outsourced labour markets, this research challenges the dominant discourse of AI as a purely technological advancement, revealing its socio-economic implications. The findings emphasize the need for ethical AI development, greater transparency in labour practices, and structural reforms to protect workers from exploitation. Without intervention, AI risks deepening global disparities, concentrating wealth and power while reinforcing systemic inequalities in the digital economy.