# AI-Assisted Systematic Review: Humans Still Need to Review All Abstracts for Inclusion

**Authors:** Hyelin Sung, Deyana Altahsh, Scott Garrison

PMC · DOI: 10.2196/82896 · JMIR Formative Research · 2026-03-19

## TL;DR

AI can help prioritize abstracts for systematic reviews, but humans still need to review all abstracts to ensure nothing important is missed.

## Contribution

Demonstrates that AI tools can assist in ranking abstracts but cannot fully replace human judgment in systematic reviews.

## Key findings

- General-purpose and fine-tuned AI models ranked abstracts for inclusion in Cochrane reviews.
- Some relevant studies were not highly ranked by AI, requiring manual review.
- Human oversight remains essential for accurate systematic reviews.

## Abstract

Although a general purpose (GPT-5), and a fine-tuned (ASReviewLab) artificial intelligence were able to rank abstracts for likely inclusion in a variety of Cochrane systematic reviews, some actually included studies were not highly ranked, necessitating human review of all abstracts.

## Full-text entities

- **Species:** Homo sapiens (human, species) [taxon 9606]

## Full text

_Full body text omitted from this summary view._ Fetch the complete paper as Markdown: https://tomesphere.com/paper/PMC13002160/full.md

## Figures

2 figures with captions in the complete paper: https://tomesphere.com/paper/PMC13002160/full.md

## References

5 references — full list in the complete paper: https://tomesphere.com/paper/PMC13002160/full.md

---
Source: https://tomesphere.com/paper/PMC13002160