Journals Face AI Submission Surge as Boston Review Sees Drafts Triple Since 2023

3 articles · Updated · The Boston Globe · Aug 2

Boston Review said unsolicited submissions in the first half of 2026 were triple the level of the same period in 2023, forcing its small staff to make faster judgments as suspected AI-written drafts pile up.
MIT Press journals report a significant uptick in suspected AI-assisted papers, including cases of one author sending as many as five papers in a day, while each submission can cost publishers $10 to $32 to process.
The New England Journal of Medicine allows disclosed AI help for clearer presentation of authentic research, especially from non-native English speakers, but rejects machine-generated opinion and is testing AI detectors for letters and essays.
Editors are also tightening policies after broader AI-writing scandals and image manipulation problems, including a May NEJM retraction over an AI-altered medical image.
Across literary, legal and scientific publishing, outlets are converging on disclosure rules and heavier review as they try to preserve human judgment while still using AI for tasks like fact-checking and translation.