This study investigates the impact of artificial intelligence (AI) involvement on the acceptance, perceived sincerity, and emotional richness of text-based communication. Building on Towne's (2024) framework, which distinguishes between AI-Assisted Communication (AI-AC) and AI-Dominated Communication (AI-DC), the study explores how these methods affect human perceptions of message quality. Two comprehensive studies were conducted using participants from the Credamo platform. Study 1 employed a between-subjects design to compare the impacts of AI involvement and disclosure conditions, while Study 2 utilized a within-subjects design to assess changes in perceptions before and after AI involvement disclosure. The findings reveal that while AI involvement does not significantly impact message acceptance, it reduces perceived sincerity and emotional richness, particularly when AI's role is disclosed. The study also highlights demographic influences, with gender and age affecting trust in AI-generated content. These results underscore the importance of transparency in AI-mediated communication and suggest strategies for managing AI disclosure to maintain trust. The research contributes to the growing body of knowledge on AI-mediated communication by validating Towne's framework and providing practical insights for integrating AI in communication processes.