#1155: Extending CSCL Affordances for Multilingual Classrooms: Teacher Perspectives on AI-Supported Collaboration

Multilingual classrooms add layers of complexity to collaboration: teachers rarely share every student’s home language, peers struggle to build common ground, and AI translation can miss linguistics nuance. We report findings from focus groups conducted with teachers of multilingual students to gauge their view on AI-supported collaboration across three collaboration pathways: student-student, teacher-student, and student-AI. Teachers welcomed real-time subtitling/translation to support comprehension and questioning, and lightweight tools that surface peers’ ideas via transcripts and highlights for clarifications. They were skeptical of in-class 1:1 student-AI partnering, citing risks to social connection, teacher workload, and scope creep. Using Jeong & Hmelo-Silver’s seven affordances as our lens, we identify where affordances blend within pathways, and where they need rethinking for multilingual equity. We translate teacher concerns into CSCL-aligned design adaptations and contribute extended versions of affordances to guide AI-integrated CSCL that is equitable for linguistically diverse classrooms.