As AI becomes increasingly embedded in evidence synthesis, conversations about “responsible use” often focus on technical safeguards, governance frameworks or performance metrics. Yet responsibility is also shaped by the language we use, the philosophical assumptions we hold, and the narratives that circulate. This discussion, part of the series: Being responsible when using AI for evidence synthesis, will:-
- Encourage participants to critically examine the narratives, metaphors, and assumptions that frame contemporary discussions of AI use.
- Surface areas where ambiguity, oversimplification, or uncritical optimism create risks for responsible practice.
- Facilitate reflection on how the Responsible use of AI in evidence SynthEsis (RAISE) recommendations and guidance (https://osf.io/fwaud/overview) might need to evolve.
Discussions may include key nuances often overlooked in responsible AI conversations, such as ambiguity in performance expectations, the danger of false confidence, the social impact of inflated claims, or the challenge of acknowledging uncertainty without undermining innovation.
The session is open to anyone interested in the evolving field of responsible AI use for evidence synthesis.
Presenter bios
Lisa Bero, an affiliate professor at the University of Colorado and Senior Research Integrity Editor at Cochrane, is a leader in evidence synthesis, meta-research and studying commercial determinants of health, focusing on tobacco control, pharmaceutical policy, nutrition, chemicals and public health. She is internationally known for her studies on the integrity of research evidence that is used to influence health policy. She is a founding member of the Center to End Corporate Harm at the University of California, San Francisco. She was co-chair of the Cochrane Governing Board from 2014 to 2018. She studies the association of reporting bias, methodological characteristics, funding bias, conflicts of interest, and research integrity problems with the design and outcomes of research.
Biljana Macura is a Senior Research Fellow at the Stockholm Environment Institute and Editor-in-Chief of Environmental Evidence, the journal of the Collaboration for Environmental Evidence. Her research focuses on advancing evidence synthesis methodologies, including the responsible use of AI to support evidence synthesis.
Prof. James Thomas is Principal Investigator of the Evidence Reviews Facility for the Department of Health and Social Care, England, which is a large programme of policy-relevant systematic reviews with accompanying methodological development to increase their relevance and applicability for decision-making. He co-led the Cochrane ‘Project Transform’ which is implementing novel technologies and processes (including machine learning and crowdsourcing) to improve the efficiency of systematic reviews, and is co-investigator on a major Collaborative Award from Wellcome, the Human Behaviour-Change Project, led by Susan Michie (UCL), which is developing technologies and methodologies to organise, synthesise and present the literature in behavioural science.
Prof. Gerald Gartlehner is a professor for evidence-based medicine and clinical epidemiology and the co-director of Cochrane Austria. He is also a co-convenor of the joint AI Methods Group and Cochrane Rapid Review Methods Groups. With over two decades of experience in conducting evidence synthesis, Gartlehner has recently focused on evaluating the performance of LLMs for semi-automating various steps in the evidence synthesis processes.
