Have the "on-hold" durations been getting longer for arXiv submissions? [D]

2026-05-13 · Source: Machine Learning · Field: Science & Research — Artificial Intelligence & Machine Learning, Research Methodology & Innovation · Depth: Intermediate, quick

Summary

arXiv submissions, particularly in the cs.AI and cs.LG categories, are experiencing significantly longer "on-hold" durations, with some users reporting waits of up to two months in 2024, a stark increase from previous turnaround times of a couple of days. This delay is attributed to an inundation of low-effort, AI-generated papers and a proliferation of "unethical" survey papers. These problematic survey papers are criticized for not genuinely advancing sparsely researched fields but rather serving as a means for authors to boost publication counts and citations due to their ease of creation and broad appeal. arXiv has also recently banned position papers and announced severe penalties for submissions containing fake citations or AI-generated content, indicating a stricter stance on submission quality.

Key takeaway

For AI Scientists submitting to arXiv, be aware that "on-hold" durations, especially for cs.AI and cs.LG categories, can now extend to two months. You should factor these extended review times into your publication schedule and meticulously ensure your submissions, particularly survey papers, demonstrate genuine research contribution to avoid delays or rejection, given arXiv's stricter enforcement against low-effort or AI-generated content.

Key insights

arXiv is experiencing submission delays due to low-effort AI papers and "unethical" survey papers.

Principles

• Survey papers should establish foundations for sparsely researched topics.
• Low-effort submissions inflate publication counts without advancing knowledge.

In practice

• Avoid cs.AI as a main category if rapid review is critical.
• Ensure survey papers address genuinely sparse research areas.

Topics

• arXiv Submission Process
• On-Hold Durations
• AI-Generated Content
• Survey Paper Ethics
• Publication Misconduct

Best for: AI Scientist, Research Scientist

Related on AIssential

• Send the arXiv AI-generated slop, get a yearlong vacation from submissions — arXiv now bans AI-generated content, imposing severe penalties on authors for non-compliance.
• No new paper under review in TMLR since May 09? [D] — TMLR's paper review process has significantly slowed, with no new submissions entering review since May 9th.
• Looking for arXiv cs endorsement — first-time submitter, paper on multi-agent LLM token optimization (Patent Pending) [D] — A first-time submitter sought arXiv cs endorsement for a patent-pending paper on multi-agent LLM token optimization.
• Am I crazy to think that the UAI authors are confusing the discussion deadline with the rebuttal deadline ? [D] — UAI's review process suffers from deadline confusion and transparency issues, hindering effective author-reviewer discussion.
• From “Math Fear” to 100 Percentile: My Mid-Sem Surprise — Consistent effort and active questioning can overcome academic challenges after a long break.
• The uncritical adoption of AI in science is alarming — we urgently need guard rails — Uncritical AI adoption in science risks reduced research quality, narrowed scope, and deskilling future scientists.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine Learning.