#1303: CLUE: A Learning-Goal Oriented AI-Enhanced STEM Inquiry System

Integrating AI into STEM classrooms risks shortcutting the productive struggle that makes inquiry-based learning effective. This session introduces CLUE, a collaborative platform that addresses this risk through rubric-centered AI personalization: rather than deploying generative AI as an open-ended chat interface, CLUE uses curriculum-tuned categorization rubrics to make student problem-solving strategies visible to both learners and teachers. In this way, CLUE’s rubric-centered approach operationalizes sensemaking for any particular curriculum. We conjecture that this design supports teacher noticing, promotes metacognitive reflection, and sustains productive struggle by shifting AI feedback away from answer-giving and toward strategy awareness and peer comparison. We will demonstrate several classroom-tested STEM curricula paired with these AI scaffolds, and attendees will participate in a live collaborative experience to see how CLUE's rubric-based approach makes AI-generated enrichment transparent and purposeful. The session will explore the design opportunities and challenges behind this methodology, invite discussion about alternative approaches, and share directions for future empirical evaluation. Attendees will leave with practical strategies for integrating AI into STEM classes and access to CLUE tools for use in their own classrooms.