Use of artificial intelligence to support evidence synthesis methods: a scoping review protocol

To date, no single review provides a clear overview of how AI, particularly LLM-based tools, is being used across the entire evidence synthesis process, including which tools are used, at which stages, and how their performance is evaluated. Evidence from individual studies and frameworks remains scattered, making it difficult for review teams to understand how AI can be used in practice. Therefore, the objective of this scoping review is to map how AI technologies are being used in the conduct of evidence syntheses. Specifically, this review will examine: 1) the stages of the evidence synthesis process where AI is applied; 2) the types of AI methods and tools reported; 3) how the performance and effectiveness of AI are evaluated; 4) the validation methods used; 5) the roles of human reviewers in AI-assisted workflows; and 6) gaps in the current evidence base.