This exploratory study investigates the impact of AI literacy training on the perception of anthropomorphism and trust in conversational AI. Anthropomorphism, the attribution of human-like characteristics to non-human entities, is increasingly impacting the way we interact with conversational AI systems. The emergence of generative AI systems such as ChatGPT, which are capable of producing highly coherent and contextually relevant responses, has amplified this phenomenon. As these systems become more powerful and prevalent, users may develop different perceptions of AI, leading to shifts in trust, reliance, and expectations, both warranted and unwarranted. This research aims to understand how AI literacy training might affect both conscious and unconscious anthropomorphic tendencies while investigating their effect on trust towards conversational AI. Forty participants were divided into a control group (N = 20) and an experimental group (N = 20). Both groups interacted with the conversational agent ChatGPT3.5 in a pre- and post-test session, between which only the experimental group received a comprehensive 3-hour AI training. The study employed a multi-faceted approach for measuring anthropomorphism, including surveys and sentiment analysis of the linguistic data obtained from the participant’s interaction with ChatGPT3.5. Self-reported data suggests the intervention of the AI literacy training did not significantly change anthropomorphism and trust levels. However, sentiment analysis using a newly created linguistic anthropomorphism scale does show a significant decrease in anthropomorphism levels after the intervention. The paper concludes with a discussion of the results and limitations, while providing relevant future research directions in this expanding domain of human-AI interaction.