The expanding application of artificial intelligence (AI) in educational systems has underscored the urgent need to redefine teachers' professional competencies. This is because the effective utilization of this technology requires not only technological knowledge but also critical thinking, ethical judgment, innovative instructional design, and the ability to responsibly evaluate the consequences of using intelligent tools. The present study aims to develop a conceptual framework for teachers' professional competencies in the age of AI—with a specific focus on critical thinking, AI ethics, and educational innovation—using a meta-synthesis approach. The study is applied in nature and qualitative in its methodological approach. Within this framework, literature concerning teacher competencies, AI literacy, technology ethics, critical thinking, and educational innovation is examined through systematic searching, screening against predetermined criteria, quality assessment, and the extraction of relevant concepts; the resulting themes are then synthesized into a conceptual framework. The proposed framework organizes teachers' professional competencies into twelve core categories: AI literacy and understanding, critical thinking, AI ethics, equity and inclusion, data literacy, human-AI interaction, intelligent instructional design, educational innovation, learning assessment and feedback, learner empowerment, continuous professional development, and the governance of AI in education. Within this framework, critical thinking serves as the basis for evaluating artificial intelligence (AI) outputs; ethics acts as a guide for responsible usage; and educational innovation provides the context for transforming technological capabilities into meaningful learning opportunities. This framework can serve as a foundation for revising teacher education programs, designing professional development courses, and formulating policies regarding the responsible use of AI in education. Nevertheless, the final validity of the proposed categories and relationships requires the rigorous execution of a meta-synthesis process, the analysis of selected studies, and the validation of the framework by experts.