This article proposes a governance model for the use of artificial intelligence (AI) in the context of personalized e-learning. The main objective is to frame the integration of AI in order to maximize its benefits, such as the personalization of learning, while minimizing the associated risks, including algorithmic bias, the protection of per-sonal data, and the transparency of decision making processes. These associated risks remain a significant challenge to be met. The approach consisted of relying on the ISACA COBIT 5 information system governance framework in order to propose a model that highlights the governance processes, objectives, and actions. The governance model presented is composed of several key elements, including transparency, fairness, data security, and continuous improvement, which are essential for effective AI management. It proposes guidelines to align AI with the strategic objectives of educational institutions, thus ensuring an ethical and responsible use of AI technologies in e-learning. This work also highlights the importance of an integrated approach to AI governance, where risks are controlled and benefits are optimized. The conclusions drawn show that, although ambitious, such a governance framework is essential to build trust among users and stakeholders, while ensuring learning that is more tailored to the individual needs of students. Finally, this suggested governance model will need to be continuously refined and adapted as the technology evolves, and calls for future research to test and validate the model in real-world contexts.