With the increasing expansion of artificial intelligence (AI) applications in public services, the necessity of developing localized regulatory frameworks for the ethical use of this technology is increasingly evident. This research aims to present a comprehensive regulatory framework for the ethical use of AI in Iran's public services, designed by considering the conditions of developing countries and localizing international experiences. This study was implemented using a mixed-methods (quantitative-qualitative) methodology. In the qualitative phase, through thematic analysis of upstream policy documents and regulatory frameworks from 15 leading countries in AI published between 2018 and 2024, as well as interviews with 28 experts and policymakers, the main dimensions and components of the framework were extracted. In the quantitative phase, a questionnaire based on the qualitative findings was designed and distributed among 345 stakeholders, including government officials, technology specialists, and civil society representatives. The findings resulted in a seven-dimensional localized regulatory framework for AI comprising: algorithmic transparency, accountability and responsibility, privacy protection, fairness and non-discrimination, human oversight, security and resilience, and technological independence and sovereignty. The implementation mechanisms of this framework were explained at three levels: hard regulation (mandatory laws and regulations), soft regulation (guidelines and ethical codes), and self-regulation (internal organizational mechanisms). Model validation was conducted using confirmatory factor analysis. This research, by presenting the first comprehensive and localized AI regulatory framework for Iran, can serve as the foundation for formulating national laws, regulations, and standards in the Islamic Republic of Iran.