ETHICS IN AI DEVELOPMENT: SOFTWARE ENGINEERS AS GATEKEEPERS OF AI DECISION-MAKING THROUGH RULES AND DATA

Artificial Intelligence (AI) systems are heavily dependent on the knowledge of software developers for creating rules, managing data, and organizing algorithms. This paper examines the crucial impact that software engineers have on the development of AI decision-making processes through the creation of rules, pre-processing of training data, and implementation of frameworks that ultimately affect model behavior. This research focuses on the effects of human-designed logic on the performance, fairness, and understandability of artificial intelligence, highlighting the interaction between automated decision-making processes and the limitations imposed by the developers involved. We also investigate the ethical implications that result from biases inherent in human-created guidelines and data collections, underscoring the necessity for accountable AI design methodologies. This study's findings are based on case studies and empirical analysis, revealing that AI systems are not entirely autonomous but rather an expression of software engineers' decision-making, expertise, and personal predispositions. We suggest implementing guidelines for software developers to minimize bias, improve the transparency of AI systems, and guarantee the ethical use of artificial intelligence in deployment. The significance of software engineers in creating dependable and trustworthy AI systems is highlighted by this research.