Modality or Authenticity?

The introduction of generative artificial intelligence (AI) tools such as ChatGPT has opened up new possibilities for language teaching, including its ability to generate texts on different topics and for different proficiency levels. While potentially helpful scaffolds for developing interpretive language skills, AI-generated texts have not yet been evaluated with respect to their impact on language learner engagement. Similarly, although research and policies cite the importance of using authentic texts in the classroom, few studies have examined the role of authentic versus “inauthentic” (e.g., AI-generated) texts in engaging learners. Therefore, in this mixed-methods, action research study we assess learner engagement around authentic and AI-generated texts in two intermediate L2 French secondary classrooms in the United States. Using data from a strategic classroom observation protocol and discussions with learners, we discover that modality, rather than text type, determined the level of engagement among these students, with significant differences found between written and audiovisual text types, rather than between AI-generated and authentic texts. Given these findings, we explore the notion of “authenticity” with regard to AI-generated texts, and discuss implications for learner engagement and the use of these texts in the language classroom.