A voice interaction system capable of appropriately handling a situation to effectively prevent a response error from occurring. A speech acquisition unit 102 acquires user speech. A feature extraction unit 104 extracts a feature of the acquired user speech. A response determination unit 120 determines a response corresponding to the extracted feature using any one of a plurality of learning models. A response execution unit 130 performs control in order to execute the determined response. A response error determination unit 140 determines whether the executed response is an error. A learning model selection unit 150 selects the learning model from the plurality of learning models stored in a learning model database 160 according to a result of the determination by the response error determination unit 140. The response determination unit 120 determines the response using the selected learning model.