How to use AI for a literature review from How to Do Your Literature Review by Gary Thomas

The emergence of AI tools for literature reviews has transformed the research landscape for both professionals and students. These technologies can swiftly scan millions of papers, deliver concise summaries, link related works, and visually map scholarly connections, making the process of gathering information more efficient than ever before. Their ability to unearth hidden or overlooked studies and to facilitate ‘snowballing’—tracing references backward and forward—means researchers can access a broader swathe of the literature, often across multiple databases simultaneously. This reduces the risk of missing important material that might not appear in subject-specific searches.

However, there are clear limitations to relying too heavily on these AI-powered aids. As highlighted in ‘How to Do Your Literature Review’, while these aids excel at compiling and organising information, they often fall short in generating the nuanced analysis and narrative required for a high-quality literature review. The human element—what might be called human intelligence (HI)—remains vital. A strong literature review doesn’t simply list sources; it interprets, compares, and synthesises them. It weaves ideas together, introduces contrasting perspectives, and offers tentative conclusions based on a balance of opinion. This narrative-making and critical evaluation are hallmarks of human scholarship, and they remain beyond the current reach of AI.

Moreover, AI tools can occasionally overlook key connections or miss obvious sources, particularly if unconventional terminology or keywords are used. As a result, while AI-based reviewing tools are invaluable for expanding the breadth and depth of literature searches, they are not a substitute for the interpretive and analytical work that only a human reviewer can provide. In summary, AI should be seen as a powerful assistant, not a replacement, and the best literature reviews will continue to rely on the narrative and critical skills of the researcher. Read more about how to use AI in a literature review in How to Do Your Literature by Gary Thomas.