HAVEN: Cooperative AI System for Trauma-Sensitive, Culturally Grounded Educational Content Generation in Low-Resource Humanitarian Settings

Ying Tang, Argya Hanisi, Tia Dwi Setiani, Aura Irfani Salsabila, Inria Astari Zahra

Keywords: human-AI collaboration, cultural grounding, educational technology, human-in-the-loop systems, trauma-informed education, cooperative AI, low-resource learning environments, AI-assisted content creation

TL;DR: A cooperative AI system that helps volunteers create trauma-sensitive, culturally grounded educational materials in low-resource conflict settings.

Abstract:

Pedagogical tools for learners living in conflict zones remain scarce, leaving volunteer educators to shoulder substantial cognitive and emotional burdens while creating trauma-sensitive, culturally grounded teaching materials under severe
resource constraints. Volunteer educators conducting unpaid virtual sessions without institutional support face high
cognitive and emotional load while creating trauma-sensitive, culturally grounded teaching materials. To tackle this, we developed \textsc{Haven}, a cooperative AI system specialized with a cultural grounding agent that assists volunteers with teaching material creation. Indonesian cultural heritage is used as an intercultural pedagogical bridge, grounding English lessons in culturally rich content without requiring volunteers to simulate learners' own heritage traditions. Our evaluation trials suggest that HAVEN’s cultural grounding mechanism changes the kinds of image-related errors that volunteer educators notice. These findings underscore that a human-in-the-selection loop is not a provisional measure but an essential check on AI image generation in non-Western heritage contexts.

Track: Track 1: ML Research Addressing Challenges Faced by Muslim Communities

Email Sharing: We authorize the sharing of all author emails with Program Chairs.

Data Release: We authorize the release of our submission and author names to the public in the event of acceptance.

Non Archival Confirmation: I understand that submissions to MusIML are non-archival and can be submitted to other venues.

Submission Number: 34