A systematic literature review (SLR) on the adoption of artificial intelligence-assisted SLRS: implications for health technology assessments

OBJECTIVES: Systematic literature reviews (SLRs) are essential for evidence synthesis in healthcare decision making, including health technology assessment (HTA), but their time and resource demands are substantial. Artificial intelligence (AI) may enhance efficiency of conducting SLRs, but its acceptance by HTA bodies remains underexplored. This SLR quantifies published health-related SLRs reporting AI use, identifies AI tools used at each SLR stage, and evaluates HTA guidance on AI in evidence synthesis. METHODS: We searched Embase, Medline, and the Cochrane Library (up to 9 September 2025), supplemented by hand searches and reviews of HTA agency websites. Titles and abstracts were screened in Rayyan by a single reviewer, with full-text review confirming eligibility. Data were extracted and synthesized narratively along key themes. RESULTS: In total, 112 studies covering 111 unique SLRs were identified, reporting 134 implementations of 45 unique AI tools (29 publicly available; 16 custom-built). AI use has risen since 2013 and was most frequently applied during title and abstract screening (88 of the 134 implementations). Human oversight remained essential, with no fully autonomous AI reported. Three HTA agencies (CDA-AMC, IQWiG, NICE), EUnetHTA, JBI and Cochrane have provided guidance, indicating the formal integration of AI into HTA processes. CONCLUSIONS: This SLR provides a quantitative overview of AI use in health-related SLRs and current HTA guidance. These findings may inform development of clearer methodological recommendations and support integration of AI-assisted evidence synthesis in HTA submissions. Further research and policy development are needed to optimize its role in evidence synthesis and healthcare decision making.