EduCompass AI

## Reto Generative artificial intelligence is rapidly transforming education, creating new opportunities for learning while introducing challenges related to academic integrity, governance, privacy, and responsible use. Educational institutions are increasingly expected to develop AI policies that balance innovation with appropriate safeguards, yet many lack the time, expertise, or governance frameworks needed to create policies tailored to their specific context. EduCompass AI was developed to address this challenge. Rather than generating a generic policy, the project uses a human-AI collaborative approach to guide institutions through a structured decision-making process that considers educational values, governance priorities, legal obligations, and acceptable AI use. The result is a customized draft policy designed to support institutional discussion, review, and refinement, while keeping human judgment at the center of policy development. ## Proceso The goal of this project is to support educational institutions in developing responsible AI use policies through a structured, human-led process. Rather than generating a generic template, the agent guides users through questions that clarify institutional context, educational values, governance needs, compliance considerations, risk tolerance, and intended AI use cases. The project aims to help institutions distinguish between beneficial AI use and uses that may undermine learning, academic integrity, privacy, or trust, critical for both students and faculty. It also aims to make policy development more practical by prompting users to consider disclosure expectations, enforcement feasibility, stakeholder responsibilities, review cadence, and next steps for institutional approval. EduCompass AI is designed to produce a draft policy that serves as a starting point for review, discussion, and adaptation. Its purpose is not to replace institutional judgment, legal review, or faculty governance. Instead, the project explores how AI can support responsible governance design by helping institutions move from uncertainty to structured decision-making while supporting responsible AI use in education. ## Resultado Developing EduCompass AI reinforced that responsible AI governance is not simply a question of whether AI should be allowed or restricted. The greater challenge is helping institutions define what learning they want to protect, where AI can enhance that learning, and where its use may undermine it. Effective policies must do more than establish rules; they must help a community understand the purpose behind those rules. This project also demonstrated that policy development is fundamentally a design problem. Educational institutions need more than better policy language; they need better decision-making processes. By structuring EduCompass AI as a guided, human-AI collaborative workflow rather than a one-step policy generator, the project encourages institutions to reflect on their values, governance responsibilities, and educational priorities before policy language is created. Ultimately, the value of EduCompass AI is not that it automates policy creation, but that it supports a more informed and structured path toward human judgment. It demonstrates how AI can be used not only as the subject of governance, but also as a tool for governance design. As generative AI continues to evolve, educational institutions will need policies that are adaptable, enforceable, and grounded in their own educational values. EduCompass AI offers one model for supporting that work by helping institutions ask better questions, make clearer decisions, and develop policies that reflect both the opportunities and the responsibilities that accompany AI.