Opportunities, challenges and risks of using artificial intelligence for evidence synthesis

‍Title: Opportunities, challenges and risks of using artificial intelligence for evidence synthesis

BMJ Evidence-Based Medicine

Authors: Waldemar Siemens, Erik von Elm, Harald Binder, Daniel Böhringer, Angelika Eisele-Metzger, Gerald Gartlehner, Piet Hanegraaf, Maria-Inti Metzendorf, Jacob-Jan Mosselman, Artur Nowak, Riaz Qureshi, James Thomas, Siw Waffenschmidt, Valérie Labonté^‍, Joerg J Meerpohl

What is the report about?

The report explores the opportunities, risks, and challenges associated with using Artificial Intelligence, particularly large language models (LLMs), in systematic reviews. It draws insights from the Methods Forum of Cochrane Germany, which was held in June 2024 in Freiburg. 

How did Laser AI help?

Laser AI was recognized as one of the AI-powered tools used to enhance systematic review processes. In addition, Artur Nowak, co-founder of Evidence Prime, attended the Methods Forum of Cochrane Germany to discuss the role of AI in systematic reviews.

What were the results?

The report highlights AI's potential and current limitations in systematic reviews while highlighting its promise in tasks like screening and data extraction. However, more work is needed to refine AI models, build researcher capacity, and engage stakeholders to ensure AI is used responsibly in evidence synthesis.

Citation: Siemens W, von Elm E, Binder H, Böhringer D, Eisele-Metzger A, Gartlehner G, Hanegraaf P, Metzendorf MI, Mosselman JJ, Nowak A, Qureshi R, Thomas J, Waffenschmidt S, Labonté V, Meerpohl JJ. Opportunities, challenges and risks of using artificial intelligence for evidence synthesis. BMJ Evid Based Med. 2025 Jan 9:bmjebm-2024-113320. doi: 10.1136/bmjebm-2024-113320. Epub ahead of print. PMID: 39788693.