International Research Journal of Education and Technology

To sum up, this paper explores the AI Regulatory Sandbox (RS) to be used as a "living lab, " shifting governance from a static compliance check to an iterative co-learning process. By analyzing sandboxes, we can see how legal derogations are provided without any compromise on the basic safety standards that are set in frameworks such as the EU AI Act. In the end, we find sandboxes reduce "Innovation Lag." In the case of both "Agentic AI" and high-risk health diagnostics, the regulatory cycle isn't keeping pace with their growing emergence.