AI-generated content is now facing strict scrutiny at the arXiv preprint server. Any submissions deemed inappropriate will result in a one-year ban and a permanent requirement for peer review before publication.

The move comes after instances of fake citations, unedited prompt responses, and nonsensical diagrams slipping past editors and peer reviewers. A key figure in these developments is Thomas Dietterich, an emeritus professor at Oregon State University and a member of arXiv's editorial advisory council and moderation team.

Dietterich highlighted that submissions to the arXiv must adhere to appropriate scholarly communication standards, including properly prepared sections, figures, tables, and references. He stressed the importance of scrupulousness and careful preparation in scientific publishing.

While AI may soon be enjoying a yearlong vacation from submissions, it’s clear that human oversight is still essential for maintaining academic integrity. The new policy aims to ensure that all published content meets the high standards expected in the scientific community.