Suitability of Chinese GenAI Platforms for Early Childhood Education: A Multifaceted Evaluation

Generative artificial intelligence (GenAI) offers transformative potential for early childhood education (ECE), yet concerns remain regarding its suitability and ethical implications for young children. This study evaluated 10 freely available Chinese GenAI platforms for suitability in ECE, considering technical performance, pedagogical adaptability, and ethical/safety considerations. Using mixed methods (quantitative scoring and qualitative content analysis), platforms were assessed on multimodal support, response speed, teaching content generation, activity organization, personalization, ethical compliance, content appropriateness, and algorithmic fairness. Results showed significant variability in performance, with some platforms excelling in technical aspects while others demonstrated stronger pedagogical adaptability. Doubao and iFlytek Spark demonstrated strong technical performance, while DeepSeek excelled in pedagogical adaptability. However, all platforms presented areas for improvement, particularly in multimodal support and the transparency of ethical guidelines. This research offers a novel framework for evaluating GenAI in ECE, which allows for a more nuanced and context-specific assessment compared to existing frameworks. This framework highlights the need for specialized ECE datasets, algorithmic transparency, and robust ethical guidelines to protect young learners. Findings provide practical guidance for educators, developers, and policymakers navigating GenAI's potential in early childhood development.