AI BriefingarXiv Imposes 1-Year Ban on LLM Papers
Key point
arXiv has clarified a 1-year ban for papers containing unverified LLM-generated errors.
Details
Thomas G. Dietterich, writing as an arXiv cs.LG moderator, stated that even content created by generative AI remains the ultimate responsibility of the author.
If there is clear evidence that unverified LLM output was included in a paper and the author did not check it, arXiv considers the entire manuscript untrustworthy.
- Penalty: 1-year arXiv ban
- Subsequent submissions: must first be accepted at a reputable peer-reviewed venue
- Examples: fabricated references, drafts with leftover LLM meta-commentary, tables where experimental figures are labeled 'illustrative'
The key issue is not the use of generative AI itself, but whether results and citations were submitted without verification.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Toss Doctor V2 Released: Maximizing QA Efficiency with LLM-based Automated Testing and Self-healing