The use of artificial intelligence (AI) has created opportunities for designing teaching and learning environments and improving teachers’ performance. This study aimed to develop a conceptual framework for an AI-based collaborative learning environment, emphasizing elementary school teachers’ performance. A qualitative approach and inductive content analysis were employed. Scientific documents published between 2015 and 2026 from domestic and international databases. Following inclusion and exclusion criteria, 60 relevant scientific documents were selected. To achieve theoretical saturation, semi-structured interviews were conducted with 12 experts, leading to identification of framework elements. For internal validation, a researcher-made questionnaire was administered to 30 participants to evaluate the framework regarding simplicity, innovation and novelty, relevance, necessity, importance, and comprehensiveness. Data were analyzed using frequency, mean, standard deviation, the Content Validity Ratio (CVR), Content Validity Index (CVI), and overall appropriateness index. Coding yielded 53 open codes, 18 axial codes, and five selective codes: collaborative learning design, learning support and facilitation, assessment, technological ecosystem, and ethics and data governance. Quantitative findings indicated an overall appropriateness index of 0.89 and a satisfaction mean of 4.33. The results indicated that the identified components can provide a conceptual foundation for an AI-based collaborative learning environment aimed at improving elementary school teachers’ performance.