arXiv has introduced new limits on the number of papers researchers can submit, following a sharp rise in submissions that it says is being driven in part by the growing use of AI tools.
The open-access repository received a record 40,363 submissions in September 2026 – almost double the 20,569 received in the same month in 2024. The surge also generated almost 9,000 support tickets for arXiv staff and moderators.
Under the new policy, which came into effect on 1 October, authors are limited to two submissions per calendar month, with no more than three active submissions at any one time.
arXiv described the measure as a stopgap while it considers how increasingly sophisticated AI tools are changing research and scholarly communication: “As of October 1st, 2026, arXiv is instituting an updated rate limiting policy across all submitters and categories as a stopgap while we determine what may be the new best practice for authors employing more and more advanced AI tools, and to give us time to improve our moderation tools and procedures accordingly.”
The scale of growth has accelerated significantly. In September 2016, arXiv received 9,869 submissions, rising to 20,569 in September 2024. This September, arXiv received 40,363 submissions. Growth has been particularly pronounced in some areas. According to arXiv, submissions to its Computer Science Artificial Intelligence category, cs.AI, have increased more than sixfold over the past two years, while submissions in other categories have doubled over the same period.
arXiv said advanced AI tools were contributing to the increase by making some aspects of conducting and communicating research faster and easier.
Its policies allow researchers to use AI tools in support of their work, provided their use is disclosed and submissions meet arXiv’s standards of scholarly interest and advance their field.
However, the repository said an increasing number of submissions were failing to meet those requirements. Moderators have reported growth in “thin” papers with a narrow scope and so-called “salami” papers, in which a single piece of research is divided into several smaller submissions. arXiv also said it had seen a marked increase in dense, AI-written papers.
The result, it said, is that authors can use AI tools to produce and submit large numbers of low-value papers much more easily than previously possible.
Thomas Dietterich, Distinguished Professor Emeritus at Oregon State University and Chair of the arXiv Editorial Advisory Council, said: “The heart of arXiv’s process for detecting and rejecting low-quality papers is our many volunteer moderators. We are so grateful that they donate their time and expertise every day. However, a relatively small proportion of authors are submitting a large number of low-quality papers and consuming a disproportionate fraction of the moderators’ time. This is unfair to authors who continue to submit quality papers – their papers can be delayed for days or weeks as a result.
“arXiv now limits submitters to up to two submissions per calendar month, with a limit of three total active submissions at any given time. This policy update is being put into effect to equitably distribute our volunteer moderators’ time across arXiv authors, and protect the arXiv corpus from a sharp increase in inappropriate submissions.”
The organisation said the new limits are an interim measure while it considers longer-term best practice for authors using increasingly advanced AI tools and works to improve its moderation tools and procedures.