AI Alignment as Pedagogical Gatekeeping: How LLMs Mediate Sensitive Intercultural Dialogue in CFL Classrooms

The repository includes the complete prompt set, the data collection and coding matrix, the intercoder reliability and repeat-testing records, the anonymized response-level teacher-rating matrix and aggregate evaluation summaries after D3 member checking, the anonymized teacher-evaluation instrument, and the D3-checked teacher-evaluation report, together with a README file describing the contents and structure of the dataset. These materials are also provided as supplementary files with the manuscript.