Öz This study focuses on the semantic prosodies of the word “robot” from words that colligates it in data of the spoken form. The data is collected from a lecturer‟s talk discussing two topics which are about man and machines in perfect harmony and the effective temperature of workplaces. For annotation, UCREL CLAWS5 Tagset is used, with Tagset C5 to select output style of horizontal. The design of corpus used is by ICE. It reveals that more positive semantic prosodies on the word “robot” are presented in the data compared to negative, with 52 occurrences discovered for positive (94,5%) and 3 occurrences discovered for negative (5,5%). Words mostly collocated with “robot” in the data are service with 8 collocations, machines with 20 collocations, surgical system with 15 collocations and intelligence with 13 collocations.