MIRACLE multi-agent AI beats generic GPT for student collaboration

Effective collaboration requires Socially Shared Regulation (SSRL) — the ability to plan, monitor, and reflect together — but students often lack these skills. Researchers Shuang Li, Haiyang Xin, and colleagues introduced MIRACLE (Multi-Agent Intelligent Regulation to Advance Collaborative Learning Environment), a specialized multi-agent AI system designed to orchestrate metacognitive regulation and proactively provide emotional and motivational support. Unlike generic chatbots, MIRACLE operates via multiple intelligent agents that coordinate to scaffold the collaborative process in real time.

In a quasi-experimental study with 90 fifth-grade students, the experimental group (n=42) used a collaborative platform called CocoNote equipped with MIRACLE, while the control group (n=48) used the same platform with a general GPT assistant. Quantitative results showed the MIRACLE group achieved significant gains across all SSRL phases — Planning, Monitoring, and Reflection — and produced higher-quality collaborative artifacts. Qualitative feedback confirmed that students perceived MIRACLE as an effective facilitator for cognitive, regulatory, and emotional support.