Employee turnover remains one of the most costly and destabilizing challenges fororganizations. For small and medium-sized enterprises (SMEs), the loss of even a single skilled employee can disrupt operations and client relationships, magnifying the need for effective retention strategies. Traditional tools such as engagement surveys or exit interviews provide only delayed and incomplete insights into workforce sentiment. Recent advances in artificial intelligence (AI) and natural language processing (NLP) have enabled sentiment analysis of employee communication. These tools promise to identify patterns of dissatisfaction earlier, allowing managers to intervene before disengagement leads to resignations. At the same time, the use of AI in human resource management raises ethical and legal concerns, particularly regarding bias, transparency, and employeeprivacy. These issues are especially critical for SMEs, which often lack the resources, governance structures, and compliance expertise of larger organizations. This thesis investigates the opportunities and ethical challenges of applying AI-supported sentiment analysis in human resource management, with a focus on SMEs. Adopting a secondary research design, it synthesizes findings from academic studies, industry reports, and regulatory frameworks. The analysis shows that sentiment tools can strengthen employee voice, provide early warning of turnover risks, and elevate HR to a more strategic role. However, the study also highlights significant risks: algorithmic bias, opacity of commercial tools, and trust erosion if employees perceive monitoring as surveillance. The results emphasize that SMEs face both the strongest need for predictive retention tools and the greatest barriers to their responsible adoption. The thesis concludes by outliningpractical safeguards, including transparency measures, data protection impactassessments, and simplified governance practices, to balance the benefits of AI-driven sentiment analysis with the ethical and legal obligations of workplace fairness.
