As artificial intelligence (AI) is rapidly integrated into evidence synthesis workflows—spanning literature retrieval and deduplication, screening, risk-of-bias assessment, data extraction and structuring, and drafting and updating evidence products. While AI may improve efficiency, it can also compromise methodological rigor through non-reproducible outputs, systematic false exclusions, distorted bias judgments, extraction errors, and inappropriate summarization. Ensuring traceability, accountability, and auditability while leveraging efficiency gains has therefore become a central challenge for the evidence synthesis community. In November 2025, four leading evidence organizations—Cochrane, the Campbell Collaboration, the Joanna Briggs Institute (JBI), and the Collaboration for Environmental Evidence (CEE)—issued a joint statement on the responsible use of AI in evidence synthesis and endorsed the responsible use of AI in evidence synthesis (RAISE) framework. The joint statement articulates core positions, including human accountability, transparent disclosure, and human oversight, whereas RAISE operationalizes these principles through actionable recommendations covering roles and responsibilities within the evidence ecosystem, requirements for building and evaluating AI tools, and ethical considerations for selecting and using AI in specific projects. This paper interprets the key elements of the joint statement and the RAISE framework, presents a scenario-based view of risk control in major stages of evidence synthesis, and discusses practical implications for Chinese-language contexts, including differences in Chinese-language text forms, heterogeneity of traditional Chinese medicine evidence, policy and journal disclosure requirements, and constraints related to cost and accessibility. The analysis aims to provide a structured reference for the responsible, locally applicable use of AI in evidence synthesis in China.