arXiv Caps Monthly Submissions at Two as AI Paper Flood Overwhelms Moderators
The preprint server arXiv will limit authors to two submissions per month starting October 2026, after monthly submissions doubled in two years and the cs.AI category grew more than sixfold, overwhelming volunteer moderators.
arXiv, the preprint server that has become the default home for new research in artificial intelligence, will cap submissions at two per person per month beginning in October 2026. The restriction is a direct response to a surge in submissions that has strained the platform's volunteer moderation system.
Monthly submissions to arXiv have doubled over the past two years. In the cs.AI category, which covers artificial intelligence, the growth has been far steeper: submissions have increased more than sixfold in the same period. The scale of the increase has made it increasingly difficult for moderators to review new papers before they appear publicly.
According to arXiv, a small fraction of authors are responsible for flooding the platform with low-quality papers. These submissions overwhelm the volunteer moderators who check that papers meet basic standards of scope and presentation. The new cap is intended to curb that behavior without shutting out legitimate researchers.
The two-per-month limit applies per person, not per research group or institution. That means a single author can still submit two papers each month, but cannot push through a larger volume of work. The policy takes effect in October 2026, giving the research community roughly a year to adjust to the new rules.
arXiv has long relied on volunteer moderators drawn from the research community. These moderators screen submissions for relevance, quality, and adherence to the platform's guidelines. As the volume of papers has grown, the workload has become unsustainable, particularly in fast-moving fields like AI where the number of new preprints has exploded.
The cap is one of the most significant changes to arXiv's submission policy in years. It reflects a broader tension in scientific publishing: preprint servers were designed to accelerate the sharing of research, but the same openness that makes them valuable also makes them vulnerable to abuse. Low-quality or spam submissions can drown out legitimate work and consume the time of unpaid moderators.
For researchers in artificial intelligence, the new limit could change how they plan their publication strategies. Some may choose to combine results into fewer papers, while others may look to alternative platforms for rapid dissemination. The policy does not affect the ability to read or cite papers on arXiv, only the rate at which new submissions can be posted.
arXiv has not announced additional measures beyond the submission cap. The platform says the change is necessary to preserve the quality and reliability of its service. Whether the cap will be enough to relieve the pressure on moderators remains to be seen, but it marks a clear acknowledgment that the current pace of submissions is not sustainable.
7
