HeadlinesBriefing favicon HeadlinesBriefing.com

Peer Review Overwhelmed in AI Era

Ars Technica •
×

As research and AI‑assisted papers surge, volunteer reviewers struggle to keep up. Jason Semprini, a health economist at Des Moines University, found that HPV mandates barely reduce cervical cancer rates. His manuscript was rejected after a single reviewer misread the study’s focus, highlighting reviewer fatigue.

Across disciplines, publication counts climb. Scopus and Web of Science grow at 5.6 percent per year, and researchers devote a collective 15,000 years of work to review each year—work that would cost US reviewers about $1.5 billion if paid. Editors now email dozens to find one reviewer.

Peer review is only ~50 years old. In the 1970s, the National Science Foundation used external reviewers to justify federal funding, turning peer review into a cornerstone of science. Historians like Melinda Baldwin note that legitimacy, not tradition, cemented the practice.

Today, AI makes writing easier but also floods journals, while paper mills threaten quality. Researchers increasingly publish through informal blogs, questioning the survival of peer review. The system feels ‘extremely protracted and painful,’ as microbiologist Sebastian Lourido describes it.