{"id":454954,"date":"2026-05-18T14:05:31","date_gmt":"2026-05-18T14:05:31","guid":{"rendered":"https:\/\/www.newsbeep.com\/ie\/454954\/"},"modified":"2026-05-18T14:05:31","modified_gmt":"2026-05-18T14:05:31","slug":"arxiv-to-ban-authors-for-a-year-over-unchecked-ai-use-in-research-papers","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/ie\/454954\/","title":{"rendered":"arXiv to Ban Authors for a Year Over Unchecked AI Use in Research Papers"},"content":{"rendered":"


\n
\n
\n <\/p>\n

The new policy reinforces that authors remain fully responsible for the integrity of their work, even if parts of it were AI-generated.<\/p>\n


\n <\/p>\n

MITSloan ME Editorial<\/a>
\n
\n less than a minute ago <\/p>\n

<\/p>\n


\n
\n
\n
\n
\n
\n
\n Subscribe
\n<\/a>
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\nShare<\/a><\/p>\n


\n Share
\n arXiv to Ban Authors for a Year Over Unchecked AI Use in Research Papers
\n
\n
\n
\n
\n Twitter<\/a>
\n
\n
\n
\n Facebook<\/a>
\n
\n
\n
\n
\n
\n
\n Whatsapp<\/a>
\n
\n
\n
\n
\n
\n
\n
\n
\n Linkedin<\/a><\/p>\n

<\/p>\n


\n \"\"\/
\n <\/p>\n

[Image: ChetanJha\/MITSMR Middle East] <\/p>\n


\n <\/p>\n

The scientific community is drawing sharper lines around accountability as the use of generative AI in science spikes. The latest signal comes from arXiv, the popular preprint open repository that has become central to how research circulates in fields such as computer science, mathematics, and physics.<\/p>\n

Although papers uploaded to arXiv are not peer-reviewed, the platform plays an outsized role in shaping scientific discourse. Researchers often use it to rapidly disseminate findings, establish precedence, and track emerging trends long before journal publication. Increasingly, however, it is also confronting the unintended consequences of widespread large language model (LLM) adoption.<\/p>\n

The repository has already introduced measures to limit low-quality or machine-generated submissions, including requiring first-time contributors to obtain endorsements from established authors. Now, arXiv moderators are signaling a tougher stance on what they see as careless or unchecked use of generative AI in research papers.<\/p>\n

In a recent policy note, Thomas Dietterich, chair of arXiv\u2019s computer science section, stated<\/a> that submissions containing \u201cincontrovertible evidence\u201d that authors failed to verify LLM-generated content would face significant penalties. Such evidence could include hallucinated citations, fabricated references, or residual prompts and comments copied directly from AI systems into manuscripts.<\/p>\n

Authors whose papers violate the rule could face a one-year ban from arXiv. After that period, any future submissions would need to be accepted first by a reputable peer-reviewed venue before appearing on the repository.<\/p>\n

The policy does not prohibit researchers from using LLMs. Instead, it reinforces a principle increasingly emphasized across academia: authors remain fully responsible for the integrity of their work, regardless of whether parts of it were generated by AI. Researchers cannot deflect blame for factual errors, plagiarism, biased language, or fabricated citations by attributing them to a chatbot.<\/p>\n

Dietterich said<\/a> the policy would operate as a \u201cone-strike\u201d system, though enforcement would include multiple layers of review. Moderators must first flag problematic submissions, after which section chairs confirm the evidence before penalties are imposed. Authors will also retain the right to appeal.<\/p>\n

The move highlights concerns about the reliability of AI-assisted scholarship. Recent peer-reviewed studies have documented<\/a> a rise in fabricated citations in biomedical research, a trend many researchers attribute to generative AI systems that confidently invent sources.<\/p>\n