1 Introduction

This section introduces the concept of learner identity construction and its significance in second language acquisition research, points out current research gaps, and emphasizes the trend of integrating AI into foreign language teaching, which leads to changes in teaching interaction modes and consequently affects learner identity construction. The study aims to explore the strategic and dynamic nature of learner identity construction in AI-integrated foreign language classrooms, using cognitive network analysis to focus on learners' role positioning and identity construction characteristics regarding themselves, teachers, and AI, in order to address two research questions.

2 Identity Construction and Superdiversity Theory

This section discusses the dynamic and complex nature of identity construction in the context of globalization and information technology development. It introduces the theory of superdiversity, emphasizing the dynamic migration, flexibility, and strategic nature of identity, as well as its contradictory, ambiguous, and negative aspects across different domains. It highlights the new perspective that superdiversity theory offers for understanding the identity construction of social subjects and its application value in researching learner identity construction in AI-integrated foreign language classrooms in the AI era.

3 Research Methodology

This chapter elaborates on the research methodology for studying the dynamic construction of learner identity in AI-integrated foreign language classrooms. The subjects are students enrolled in the compulsory course "Language and Education" for English majors at a comprehensive university in Shanghai. The course implemented AI-integrated teaching reform starting in the fall of 2024, incorporating four AI-related interactive tasks to investigate learner identity construction characteristics. Data were collected through semi-structured and retrospective interviews involving seven students. Data analysis employed cognitive network analysis, which involved establishing a coding framework, data annotation, and modeling analysis to visually present the association patterns and adjustment processes among identity elements, thereby capturing the dynamic and interconnected nature of learner identity construction.

4 Research Findings

Through cognitive network analysis, this section explores in detail the dynamic changes in learner identity construction during interactions with AI in foreign language classrooms. The findings reveal that before tasks, learners often perceive AI as a tool or assistant and construct their learner identity primarily from the teacher-student relationship, while clearly distinguishing the roles of AI and teachers and strategically consulting different sources. During tasks, learner identity construction becomes more diverse, including roles such as AI explorer and evaluator. The instrumental positioning of AI is deconstructed, incorporating it into the social relationship network, redefining human-machine relationships, with task characteristics significantly influencing identity construction. After tasks, learner roles diversify further; the instrumental nature of AI and the guiding role of teachers remain prominent, but learners also manifest identities as AI explorers and evaluators. Their identity construction evolves with experience, offering new perspectives and reflections.

5 Discussion

This chapter discusses the diversity, adaptability, contextual sensitivity, and iterative nature of learner identity construction in AI-integrated foreign language classrooms. Cognitive network analysis reveals that teachers, students, and AI assume multiple roles during interactions, and different learners construct distinct identity relationship patterns—hierarchical, cooperative, and competitive—under the same tasks. Identity construction is influenced by learners' technology usage habits, prior experiences, and task interpretation, demonstrating heterogeneity and diversity. Learners continuously adjust their identities across tasks, reflecting contextual sensitivity and adaptability. Identity construction is iterative, with new identity experiences supplementing the existing identity repertoire. AI integration makes learner identity construction more complex and dynamic. The triadic interaction model of "teacher–student–AI" provides rich identity references, accelerates the process of dynamic identity transformation, weakens identity boundaries, and enables learners to utilize AI as a technological resource, breaking free from the constraints of a singular language learner identity and achieving superdiverse practices. The analytical dimensions of superdiversity theory can be extended from language use to technology use.

6 Conclusion

This section summarizes the innovations and findings of the study, noting that cognitive network analysis reveals the diverse, adaptable, context-sensitive, and iterative characteristics of learner identity construction in AI-integrated foreign language education. It emphasizes that teachers can leverage AI to cultivate learners' diverse identities and rational attitudes toward technology, necessitating proficiency in AI technology and exploration of teaching strategies. It recommends incorporating AI literacy into teacher training. The study acknowledges limitations, such as the small sample size limiting the scope of identity labels, and suggests future research could expand the participant pool to explore more comprehensive identity construction characteristics.