Design Implications for Teacher-Configurable Generative AI Writing Tools: A Participatory Design Study with StoryBot

Abstract

As generative AI (GenAI) tools become accessible, their integration into K–12 writing presents opportunities and challenges. This study partnered with 21 primary school teachers in a participatory co-design process to develop StoryBot, a teacher-configurable GenAI story-authoring platform. Teachers identified instructional needs and learning outcomes that informed system features, including genre-specific writing scaffolds, configurable prompts, and constraints on full-text generation to preserve student voice. The platform positions educators as co-designers of AI behavior rather than passive users. We conducted a pilot writing workshop with 13 youth (ages 8–12) and a focus group with six educators to examine how the platform influenced students’ writing behaviors and classroom integration. Findings demonstrate the value of teacher co-design and configurable AI systems in aligning GenAI tools with pedagogical priorities and practices. We present design implications for teacher-configurable GenAI writing systems that support student agency, ethical safeguards, and instructional alignment in human–AI coauthoring environments.