Exploitative leadership behaviors, particularly toxic leadership continue to challenge organizations, especially in increasingly digital work environments where communication often lacks transparency. This paper examines how natural language processing (NLP) and AI can be used to detect and support ethical leadership by aligning corporate social responsibility (CSR). We introduce a custom NLP tool, powered by ChatGPT, designed to detect language cues commonly linked to toxic leadership, such as hostility, manipulation, and exploitation. Drawing on established theories related to digital communication, leadership ethics, and CSR, we present the development process of this tool and test its relevance through an employee-focused survey. The findings highlight the potential for AI to act as a practical mechanism for identifying and mitigating harmful leadership behaviors in real time. This research offers both theoretical insights and practical implications, suggesting that responsible AI can strengthen CSR initiatives and help foster ethical, psychologically safe workplace cultures.