arXiv Is Rate Limiting Submissions Because It Can’t Keep up With AI Slop
The academic publisher said submissions have doubled in the past two years and moderators are having a hard time keeping up.
arXiv, the open-access repository for preprint academic research, has implemented rate limiting on submissions due to an influx of AI-generated papers. Researchers can now submit only two papers per month and maintain a maximum of three active submissions at any given time. In a blog post announcing the change, arXiv explained that the issue stems from the ease with which AI can be used to flood their inbox, overwhelming their volunteer moderation team.
Thomas Dietderich, chair of arXiv’s editorial advisory council, praised the moderators for their contributions, but noted that a small number of authors are submitting a large volume of low-quality papers, causing delays for those who submit high-quality work. The number of submissions to arXiv has doubled in the past two years and has increased sixfold in the computer science category.
AI is contributing to this surge through various means, including researchers using large language models to analyze extensive datasets, write code, and design experiments, as well as some using AI to compose the papers themselves. ArXiv has permitted the use of AI as a research tool, as long as it is disclosed and the research adheres to their standards, but many of the recent submissions do not meet these criteria.
The surge includes thin papers with narrow scope, "salami" papers based on one study, and dense, AI-written papers. Rate limiting was already a part of arXiv's policy, but the new rules aim to distribute the workload of their volunteer moderators more equitably. This move is part of a broader struggle for the publication against AI-generated low-value content.
Less than a year ago, arXiv ceased accepting review articles and position papers in the computer science category, citing the low effort of LLM-written articles. In January, they began requiring new authors to have an endorsement from an existing user, previously only requiring an academic email address. Most recently, arXiv announced that researchers submitting AI-generated low-quality content would be banned from the site for a year.
Written by urgent.news from 404 Media's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
