TL;DR
ArXiv will impose a one-year ban on authors caught submitting AI-generated papers with clear evidence of inaccuracies or misconduct. The new policy aims to combat AI slop and preserve research quality amid rising fraudulent submissions.
ArXiv, the prominent open-access preprint repository, will ban authors for one year if their submissions contain incontrovertible evidence of AI-generated inaccuracies or misconduct, according to a statement from Thomas Dietterich, chair of the computer science section. This measure aims to address the growing problem of AI-generated research slop and maintain the integrity of scientific submissions.
On Thursday evening, Thomas Dietterich announced that ArXiv will enforce a one-year ban on authors whose papers include clearly identifiable AI-generated errors, such as hallucinated references or misleading comments from language models. The policy specifies that the ban applies after a single confirmed violation, with the possibility of appeal. The enforcement process involves moderation and confirmation by the section chair before the penalty is applied.
ArXiv’s move follows a series of measures taken since November 2025, including the suspension of computer science review articles and the requirement for first-time submitters to have endorsements, aimed at curbing AI-fueled fraudulent submissions. Recent studies, such as one by Columbia University, indicate a significant rise in fabricated citations, with one in 277 papers in early 2026 containing false references—an increase from previous years—highlighting the strain on peer review caused by AI-generated content.
Why It Matters
This development is significant because it directly targets the surge of AI-generated ‘slop’ that threatens the accuracy and reliability of scientific literature. By imposing strict penalties, ArXiv aims to deter authors from submitting low-quality or fabricated research, thereby protecting the credibility of preprints and the broader scientific community. The policy reflects growing concerns over AI’s impact on research integrity and the need for stricter oversight in open-access repositories.
XXX Detection Stick – Illicit Image and Video Scanner for Windows | Find and Wipe NSFW Content | Porn Blocker | Porn Detection USB Stick
Portable USB Solution — The XXX Detection Stick arrives ready to use straight out of the box —…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Background
Since late 2025, ArXiv has faced increasing challenges with AI-generated submissions, prompting several policy changes. The repository’s decision to ban authors for a year follows an earlier announcement that it would no longer accept certain categories of papers due to flooding with AI slop, especially in computer science. The rise of AI tools capable of producing plausible but inaccurate scientific content has exacerbated issues like fabricated citations and misleading data, straining peer review processes and undermining trust in preprints.
“If generative AI tools generate inappropriate language, plagiarized content, biased content, errors, mistakes, incorrect references, or misleading content, and that output is included in scientific works, it is the responsibility of the author(s). We have recently clarified our penalties for this.”
— Thomas Dietterich
“This change will help arXiv raise more money from a wider range of donors, which is needed to deal with the emergence of ‘AI slop.'”
— Greg Morrisett
plagiarism checker for research papers
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What Remains Unclear
It remains unclear how strictly the policy will be enforced in practice, how ‘incontrovertible evidence’ will be determined, and whether similar measures will be adopted by other repositories. Details about appeals processes and the scope of violations are still developing.
RECOLX AI Voice Recorder & Transcriber with GPT-5.2 Analysis – 30-Hour Recording, 112-Language Speech-to-Text & Auto Summary for Meetings, Lectures & Interviews,Grey
GPT-5.2 AI Transcription & Summary Turn hours of audio into clear text and concise key-point summaries with GPT-4o/5/5.2/0SS-120b,…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What’s Next
Next steps include the implementation of the ban policy, monitoring its effectiveness, and possibly expanding enforcement procedures. Further guidance on identifying AI-generated misconduct and the appeals process is expected to be clarified in the coming weeks.
Anatomy of the Organs: QuickStudy Laminated Reference Guide (Quick Study Academic)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
What constitutes incontrovertible evidence of AI-generated misconduct?
According to arXiv, evidence includes clear signs such as hallucinated references, meta-comments from language models, or fabricated data that cannot be reasonably disputed.
Can authors appeal a ban under this policy?
Yes, authors will have the opportunity to appeal, with decisions subject to review by the moderation team and section chair.
Will this policy apply to all scientific categories on arXiv?
Initially, the policy targets computer science submissions, but it may be expanded to other categories as AI-generated content issues grow.
How will arXiv verify if a submission contains AI-generated errors?
Verification involves moderation and confirmation by section chairs, based on identifiable signs such as hallucinated references or meta-comments from language models.
What impact does this have on open-access research?
This policy aims to preserve research integrity by discouraging low-quality or fraudulent AI-generated papers, but it also raises questions about the balance between openness and quality control.
