Enhancing foreign language learning through AI-driven telecollaboration

Artificial Intelligence (AI) is transforming language education through individualized practice, instant feedback, and enhanced engagement. This study investigates its integration into Technology-Enhanced Project-Based Language Learning (TEPBLL) and telecollaboration, focusing on perceptions of 48 young English learners who used a personalized version of ChatGPT-4 during a transnational project on clean water and sanitation. A qualitative thematic analysis of post-project survey responses was conducted, and theme frequencies were quantified to gauge salience. Learners reported high perceived value (90.7% positive) across language production, comprehension, collaboration, and AI/digital literacies. However, the teacher remained the preferred feedback source (69%). So, while AI offers immediacy and iteration, it cannot replace the social-emotional dimension of learning. Students also flagged challenges such as occasional misinformation, overloaded outputs, prompt-formulation, and language barriers. Findings suggest a pragmatic integration of AI as a scaffold to support teacher-led, telecollaborative TEPBLL, highlighting the need for explicit prompt literacy and mediation.