This paper explores the use of ChatGPT to provide personalized, human-like explanations for recommendations in recommender systems (RSs). The study evaluates how users perceive explanations generated by ChatGPT compared to random (but popular) recommendations. The research focuses on whether users prefer ChatGPT-generated recommendations over random ones and whether personalized explanations are more effective than generic ones, especially for unfamiliar items. The study involved 94 participants who were asked to provide preferences for movies they liked and disliked. Based on these preferences, ChatGPT generated four recommendations and two disrecommendations, along with explanations. Participants then evaluated the recommendations and explanations based on factors such as effectiveness, personalization, and persuasiveness. The results show that users tend to prefer ChatGPT-generated recommendations over random ones. Personalized explanations were perceived as more effective than generic ones, particularly for unfamiliar movies. However, users did not perceive user-based explanations as significantly more personalized than generic ones unless the recommendations were random. Additionally, explanations for unfamiliar movies were found to be more effective than those for familiar ones. The study also found that explanations generated by ChatGPT were persuasive and effective, but there was no significant difference between user-based and generic explanations in terms of persuasiveness. The findings suggest that ChatGPT can provide effective and persuasive explanations for recommendations, but further research is needed to understand the full impact of personalized explanations in RSs. The study highlights the importance of personalization in explanations and the need for further research into the effectiveness of different explanation types in RSs.This paper explores the use of ChatGPT to provide personalized, human-like explanations for recommendations in recommender systems (RSs). The study evaluates how users perceive explanations generated by ChatGPT compared to random (but popular) recommendations. The research focuses on whether users prefer ChatGPT-generated recommendations over random ones and whether personalized explanations are more effective than generic ones, especially for unfamiliar items. The study involved 94 participants who were asked to provide preferences for movies they liked and disliked. Based on these preferences, ChatGPT generated four recommendations and two disrecommendations, along with explanations. Participants then evaluated the recommendations and explanations based on factors such as effectiveness, personalization, and persuasiveness. The results show that users tend to prefer ChatGPT-generated recommendations over random ones. Personalized explanations were perceived as more effective than generic ones, particularly for unfamiliar movies. However, users did not perceive user-based explanations as significantly more personalized than generic ones unless the recommendations were random. Additionally, explanations for unfamiliar movies were found to be more effective than those for familiar ones. The study also found that explanations generated by ChatGPT were persuasive and effective, but there was no significant difference between user-based and generic explanations in terms of persuasiveness. The findings suggest that ChatGPT can provide effective and persuasive explanations for recommendations, but further research is needed to understand the full impact of personalized explanations in RSs. The study highlights the importance of personalization in explanations and the need for further research into the effectiveness of different explanation types in RSs.