What is this?
EduPolicyAI Monitor is an experimental web application designed to monitor, analyze, and visualize state-level AI-related educational policies in the United States. It integrates large language model (LLM)-powered summarization and pedagogical assessment tools to support timely understanding of policy trends.
Why did we build this?
As AI technologies rapidly evolve, so do the educational policies governing their integration. However, policymakers, educators, and researchers often face challenges keeping up with fragmented updates across different states. This tool was developed to bridge that gap by providing:
- Concise AI-generated summaries of official announcements, reports, and policy briefs
- Evidence-backed policy extracts using RAG-like prompting strategies
- Impact classification scores based on pedagogical relevance (e.g., symbolic vs. curriculum-level changes)
- Interactive visualizations of policy distribution across states
How does it work?
The platform utilizes Google’s Genkit and Gemini LLM to process and summarize public policy text sources. It generates structured outputs, including a summary, evidence quote, impact score (0–3), and pedagogical rationale. These are visualized on a U.S. heatmap and can be filtered by keywords or policy domains.
All AI responses are logged (anonymously) and optionally rated by users, enabling continuous evaluation and model improvement.
Who is this for?
- Educators seeking up-to-date insights on AI-related teaching guidelines
- Policymakers comparing policy diffusion across states
- Researchers analyzing the pedagogical framing and educational impact of AI initiatives
Academic Foundations
The pedagogical analysis is grounded in frameworks such as:
- LXD (Learning Experience Design) — including design principles from Niels Floor, Schmidt & Tawfik
- Symbolic vs. Structural Policy Typologies
- Explainable AI (XAI) and generative text transparency design
Research Context
This tool is part of an academic project exploring how LLMs can support transparent, scalable, and context-sensitive interpretation of education policy. It is designed for use in research pilot studies and will evolve as part of ongoing evaluation and iteration.
Contact & Citation
This prototype was developed by the ADDIE Lab at The University of Alabama. For research collaboration or questions, contact jmoon19@ua.edu.
If citing, please reference: Moon, J. (2025). AI-supported policy analysis tool for educational innovation. Unpublished manuscript.