Using AI-Generated Songs in Foreign Language Learning

This study investigates the impact of AI-generated songs on vocabulary acquisition and learner motivation in foreign language education. Conducted at a secondary vocational school in Prague, the research involved learners of German as a second foreign language from the third-year classes. Using a quasi-experimental design with 60 observations – 30 in the experimental conditions and 30 in the control conditions – the study compared conventional vocabulary learning methods with an innovative approach involving the AI music platform Suno. In the experimental conditions, the learners achieved a significantly higher mean post-test score (6.2) compared to the control conditions (4.2), with a statistically significant difference confirmed at p < 0.001 and a large effect size (Cohen’s d > 0.8). The survey data revealed that the learners preferred the song-based activity, citing increased engagement and enjoyment. These findings suggest that AI-generated songs can significantly enhance both vocabulary retention and student motivation, particularly in contexts requiring the acquisition of specialised terminology, which is often more challenging because of its complexity and the limited availability of learning resources.