This research paper explores the burgeoning intersection of Affective Computing and Human Resource Management (HRM), specifically focusing on how Artificial Intelligence (AI) can generate quantifiable "emotional well-being scores" for employees. As organizations pivot toward hyper-digitized, data-driven HR practices, emotional AI tools are transitioning from experimental concepts to corporate realities. These sophisticated systems analyze multi-modal data—ranging from vocal paralinguistics and micro-facial expressions to semantic writing styles and digital behavior patterns—to predict an individual's internal emotional state. This paper provides a comprehensive analysis of the benefits of proactive mental health intervention, the inherent risks of "emotional surveillance," and the profound ethical concerns regarding privacy and algorithmic bias. By synthesizing current literature and technological trends, the study evaluates the potential for these numerical scores to inform high-stakes HR decisions such as promotions, workload adjustments, and retention strategies. The findings conclude that while AI-generated emotional well-being scores offer a revolutionary path toward a more empathetic workplace, their utility is strictly contingent upon a framework of transparency, employee consent, and the preservation of human-centric empathy in decision-making processes.