Security1 publisher2 min readPublished
arXiv limits researchers to two submissions a month to protect its volunteer moderators
arXiv now limits each researcher to two submissions a calendar month and three active at once to curb a flood of AI-written papers. Moderators who once throttled heavy submitters at their own discretion now apply one fixed rule to every account.
The Watch · Security desk
Drafted by a language model from the sources cited here and checked against its claim ledger before publication. How we use AISend a correction
What happened
- Submissions in arXiv's computer science category have grown sixfold, according to the archive.
- Advisory council chair Thomas Dietterich said authors who still submit quality papers can see them delayed by days or weeks as a result of the low-quality load.
- arXiv permits authors to use AI in their research if they disclose it, but says many papers it now receives miss its standards of scholarly interest.
- The archive reports more thin papers of narrow scope, more salami papers cut from a single study, and a marked rise in dense AI-written papers.
Compiled by The WatchSomething wrong?How this is made
Why it matters
- constraint At 24 submissions a year per researcher, the ceiling binds productive authors as tightly as the heavy low-quality submitters it was written for.
- exposure Because the cap counts per researcher, it holds only while new identities stay costly; anyone who can win endorsements for extra accounts multiplies their monthly allowance.
- precedent arXiv says AI tools are flooding other repositories as well, and its fixed per-account cap gives any volunteer-reviewed intake queue a public example to copy or measure against.
arXiv's intake took eight years to double once. September submissions rose from 9,869 in 2016 to 20,569 in 2024 [7][8][20]. They nearly doubled again in the next two years, to 40,363 in September 2026 [9][21]. That month averaged about 1,345 submissions a day for a volunteer moderation team that arXiv says is overwhelmed [22][2].
Thomas Dietterich, an Oregon State University professor emeritus who chairs arXiv's editorial advisory council, put those volunteers at the center of quality control [6]. "The heart of arXiv's process for detecting and rejecting low-quality papers is our many volunteer moderators," he said [3]. He described where the strain comes from: "a relatively small proportion of authors are submitting a large number of low-quality papers and consuming a disproportionate fraction of the moderators' time." [4]
A per-researcher limit is aimed at that pattern. The new ceiling bounds how much any one researcher can put in front of a moderator in a month [1]. It counts submissions. It needs no ruling on whether a model wrote them [1]. arXiv framed the change in terms of moderator hours: "This policy update strives to fairly distribute the incredibly valuable work of our volunteer moderators who check that submissions meet these standards," the archive said [14].
The cap is the fourth control arXiv has published in a little under a year [24]. Just under a year ago it stopped accepting review articles and position papers in computer science, citing low-effort LLM-written submissions [15]. In January it began requiring new authors to get an endorsement from an existing author; an academic email address had been enough before [16]. In May it said it would ban for a year researchers who submitted what 404 Media calls AI slop [17]. 404 Media's report does not mention automated detection among those measures, or say how many authors exceed the new limits [19].
What to watch
- Moderation turnaround or rejection figures from arXiv after the cap takes effect, the first evidence of whether two a month frees enough volunteer time.
- Evidence of heavy submitters spreading papers across newly endorsed accounts, and any tightening of the January endorsement rule in response.