The disclosed embodiments illustrate methods and systems for detecting plagiarism in a conversation. The method includes receiving first input corresponding to a query from a first user in said conversation. The first input corresponds to at least a first audio signal received from said first user. The method includes receiving second input corresponding to one or more responses received from a second user in response to said query. The second input corresponds to at least a second audio signal received from said second user. Thereafter, the method includes determining a first score for one or more websites, based on a comparison between said one or more responses and content obtained from said one or more websites in response to said query. The first score is a measure of a similarity between said one or more responses and said content. The method is performed by one or more microprocessors.