A system and method for using augmented reality for assisting speech development of multiple languages includes analyzing speech recorded in an environment to detect a word used in a conversation between a first speaker and a second speaker that is in a language different from other words in the conversation, wherein a target user is in the environment with the first speaker and the second speaker, isolating an object associated with the word within an augmented reality environment of the target user located in the environment, determining a confusion level of the target user based on a use of the word in the conversation, and implementing, by the processor, an augmented reality technique based on the confusion level of the target user.