An Artificial Intelligence-Enabled Tool for Real-Time Intercultural Competence Instruction

Abstract

This paper reports on a collaborative action research project between the Defense Advanced Research Projects Agency (DARPA) and the Defense Language Institute, Washington Office (DLI-W). The project examined the utility of using an existing interpretation system, not originally created for instructional purposes, for teaching and assessing intercultural competence. The system studied was the Cross-cultural Interaction Real-time Assistant for Negotiators and Operators (CIRANO), an artificial intelligence (AI)-enabled, consecutive interpretation and cultural norm violation detection system. CIRANO extends the ability of normal digital interpretation tools by adding capabilities to identify cultural norm violations and to detect emotional cues, and provides real-time rephrasing suggestions to ease interpersonal communication. CIRANO’s utility for culturally-based foreign language instruction had not been explored. DARPA and DLI-W thus collaborated to pilot CIRANO with 19 U.S. Air Force students and four native-speaking instructors participating in a 64-week Mandarin Chinese language program. Students and instructors incorporated the tool into a scenario-based lesson. Analysis of student and instructor feedback and research assistants’ observation notes indicated that CIRANO enhanced awareness of cultural differences, supported the practical application of intercultural competence principles, and generated valuable instructional interactions. However, serious limitations included the absence of monolingual support, the lack of context-sensitive feedback, and a heavy reliance on reading screen prompts. CIRANO shows potential for classroom integration, contingent upon further refinement to align with DLI-W’s pedagogical environment, and to advance broader Department of War objectives in language and culture education.