Curie Brief
Turn on cookies to sign in
Signing in saves your progress to your Curie account. We can only do that with cookies on — turn them on to continue.
News
AI is helpful, but not there yet. A new analysis finds that AI platforms can assist with early-stage literature searches but still miss too many studies to replace traditional systematic review methods. Key gaps include poor coverage of gray literature and low overall sensitivity — sometimes capturing as little as 6% of relevant studies.
AI is reshaping how researchers find and organize scientific literature, but a new analysis out of Getafe University Hospital in Spain is urging caution. The verdict: AI tools are useful companions in the systematic review process, but they're not yet ready to replace the expertise of trained information specialists.
The core problem is what researchers call the "precision mirage" — AI platforms tend to surface highly relevant documents at the top of results, making them appear impressively accurate. But dig deeper, and the gaps become significant. These tools often miss a substantial portion of eligible studies, particularly those without a DOI, such as conference abstracts and unpublished trial registries.
The analysis recommends a hybrid approach: use AI to explore concepts, spot gaps, and flag missed citations — but keep human specialists firmly in charge of the full search strategy.
By the Numbers:
Why it matters: Systematic reviews are the backbone of evidence-based medicine. If AI tools quietly miss key studies, clinical guidelines built on incomplete evidence could have real consequences for patient care — making transparency and human oversight essential.
