A comprehensive new artificial intelligence (AI) tool has launched. It can help spot problematic predatory science journals and protect the quality of research coming out of academia. This cutting-edge tool has already flagged more than 1,000 previously unknown suspect journals. Collectively, these journals publish hundreds of thousands of articles each year, leading to fodder for serious concerns about the quality and reliability of scientific literature.
There was extensive training for the researchers. To do this, they used a dataset that included more than 12,000 high-quality journals and about 2,500 low-quality or questionable publications. This AI tool then was put to work on a dataset of 93,804 open-access journals. These journals were obtained from Unpaywall, a database that assists people in locating free versions of academic articles.
Positive report showing the AI tool’s findings shows there was a 24% false positive rate. That’s a lot—it adds up to about one out of every four scholarly, legitimate journals that are at risk of being mistakenly flagged as suspect. This rate can impose difficult barriers on the academic community. It creates more confusion between reputable journals and predatory journals.
“Our findings demonstrate AI’s potential for scalable integrity checks, while also highlighting the need to pair automated triage with expert review.” – Han Zhuang et al
The research surrounding this AI tool was published in an article by Paul Arnold, edited by Gaby Clark, and fact-checked by Robert Egan. I snagged these results from phys.org on August 29, 2025. Together, they make a tremendous addition to the continuing conversation about research integrity.
As we enter an age where the integrity of scientific literature is more closely questioned, the ability to identify suspect journals becomes a valuable asset. We know that the academic landscape is ever-changing. This AI application should be an important game-changer for researchers and institutions alike, enabling them to build greater confidence in the literature they’re relying on.