arXiv updates its practices after being inundated with AI-generated 'research papers'

According to their press release, the Computer Science (CS) category of arXiv has updated its moderation practices regarding survey articles and position papers. Before they can be considered for submission to the CS category of arXiv, survey articles and position papers must now have been accepted by a journal or conference and have undergone a successful peer review process. When submitting review articles or position papers, authors must include evidence of a successful peer review process in order for their contributions to be fully considered. Review/survey articles or position papers submitted to arXiv without this evidence are likely to be rejected and will not be published on arXiv.

This change is being implemented due to the unmanageable influx of review articles and position papers to arXiv CS. In recent years, arXiv has been inundated with articles. Generative AI/large language models have contributed to this flood by making it quick and easy to write articles, especially those that do not present new research findings. While all categories on arXiv have seen a sharp increase in submissions, this is particularly pronounced in the CS category of arXiv.

We are on LinkedIn

SCIPROM is now on LinkedIn! Follow us for more information from us, the EC and the projects we manage!

News archive