The current monolithic approach in teaching and assessing English as a foreign language (EFL) communication in Indonesia has resulted in a gap between students’ understanding of the nature of communication skills and their ability to use them correctly in their speaking, leading to anxiety, lack of self-confidence and lack of mastery vocabulary. Computational thinking (CT), a high-level approach to problem-solving and optimizing knowledge acquisition, was used to design solutions in this study using a Virtual Robot application to stimulate their speaking motivation. They are taught using a CT teaching model. To find out the impact of this method, Pre-test and post-tests were conducted. All undergraduate students showed considerable improvement in their use of vocabulary when speaking, a new level of agency and interest in learning English was evident and their dialogue confidence improved. In the field of English as a foreign language (EFL), CT is an important social mediation technique. This research is the best attempt to incorporate CT into the thinking process of students learning English with the virtual assistance of robot artificial intelligence.