In the education sector, generative AI is gaining traction across applications such as intelligent tutoring, automated feedback, content generation, assessment support, and student advice. The use of these applications can be used to expand access to learning and provide personalized learning experiences. They also introduce relevant questions, including those about the potential for inaccurate or discriminatory results, opaque algorithms, privacy issues, overreliance on automated systems and unequal outcomes. This article suggests that equity-based oversight is a socio-technical approach thaed to help guide the use of generative AI in education. They are designed in keeping with the principle of fairness, inclusion, transparency, accountability and contestability, which are not an afterthought or a part of the process after they haveIt highlights the need to engage students and educators in the design of the systems, to explain why their work is important and meaningful, to provide an understanding of the outputs from AI, to include effective human review, to conduct regular auditing of AI to ensure equity, fairness, and the prevention of harmful interactions.ul interactions. The article goes beyond the technical aspects of AI's performance to emphasize the implications for educational justice, institutional accountability, and learner protection. It provides actionable insights on creating and deploying generative systems that can improve personalization without worsening social and learning inequities.