FAIRNESS IN THE WORK BEHIND THE AI INDUSTRY: HOW ACTION-RESEARCH APPROACHES CAN BUILD BETTER LABOUR CONDITIONS

Abstract

The paper presents findings about the labor conditions on the work behind AI development, with two main goals. First, the aim is to investigate the current labor conditions in the AI industry and point out the lack of quality of outsourced jobs offered by digital labor platforms and Business Process Outsourcing companies, calling attention to the challenges faced by workers and the human costs of AI systems development. Second, it presents a global research project results on how not only to carry on international studies on the topic but also how to use an action-research approach to generate impact through public scoring companies and encourage the adoption of best practices by them. The paper’s methods are based on the project’s action-research approach and methodological framework. The project scores companhies based on principles that address the major issues that define labour relations, namely, pay, conditions, contracts, management, and representation. The paper presents the result of the assessment of cloudwork platforms conducted in 2023. In addition, it paper will examine labour relations in BPO companies in the AI supply chain using a case study of the company Sama, which is based in the United States and operates in many African countries. Among the findings are problems such as non-payment situations, lack of policies to ensure minimum wage, significant rates of unpaid labour, poor measures to promote health and safety, problems in management practices and weak collective bargaining practices.