TL;DR
ArXiv will impose a one-year ban on authors whose papers contain confirmed AI-generated errors or hallucinated references. Future submissions must be peer-reviewed before posting. The move aims to curb low-quality AI-driven research papers.
ArXiv has announced a new policy to ban authors for one year if their submissions contain incontrovertible evidence of unchecked AI-generated content, aiming to reduce low-quality papers filled with AI ‘slop.’ The policy emphasizes accountability for authors and requires future submissions to be peer-reviewed before posting.
According to Thomas Dietterich, ArXiv’s section chair of computer science, authors whose papers include clear evidence of unverified large language model (LLM) outputs—such as hallucinated references or meta-comments—will face a one-year ban from submitting to ArXiv. This move is part of a broader effort to improve research quality amid the rise of AI tools that can generate superficial or misleading content.
Dietterich clarified on X (formerly Twitter) that the policy applies only when there is incontrovertible evidence of AI misuse, which includes hallucinated references or meta-comments indicating unverified AI output. The process involves a moderator documenting the issue, followed by confirmation from the section chair before imposing the ban. Authors can appeal decisions.
Additionally, ArXiv has tightened its policies for computer science review articles, requiring that such papers be peer-reviewed and accepted at reputable conferences or journals before posting. This aims to prevent the proliferation of low-effort review articles that are often just annotated bibliographies, according to ArXiv officials.
Why It Matters
This development is significant because it signals a stricter stance by ArXiv against low-quality research increasingly influenced by AI tools. It underscores the importance of accountability in scientific publishing and may influence how researchers approach AI-generated content, emphasizing verification and peer review. The policy could set a precedent for other preprint servers and academic platforms seeking to maintain research integrity amid rapid AI advancements.
AI Programming Made Practical: A Step-by-Step Guide to Building AI-Powered Applications, Writing Better Code Faster, and Using Modern AI Tools with Confidence
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Background
Over the past year, ArXiv has taken steps to address AI-generated ‘slop,’ including restricting certain review articles to peer-reviewed venues. The rise of large language models has made it easier for researchers to churn out superficial papers, prompting platforms like ArXiv to implement stricter policies. This move follows broader concerns within the academic community about the quality and trustworthiness of AI-augmented research outputs.
Previously, ArXiv highlighted issues with review articles that lack substantial discussion of research challenges, often reduced to bibliographies. The new policy on AI-generated content builds on these efforts to uphold scientific standards and prevent the dissemination of misleading or unverified information.
“If a submission contains incontrovertible evidence that the authors did not check the results of LLM generation, this means we can’t trust anything in the paper.”
— Thomas Dietterich
“Our Code of Conduct states that by signing your name as an author of a paper, each author takes full responsibility for all its contents, irrespective of how the contents were generated.”
— Thomas Dietterich
plagiarism detection software for research papers
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What Remains Unclear
It is not yet clear how frequently ArXiv will enforce these bans or how many papers might be affected in the near term. The criteria for incontrovertible evidence and the appeals process are still being clarified, and the policy’s effectiveness in reducing low-quality AI-generated papers remains to be seen.
Debugging the Code Review: How Top Engineers Improve Their Craft Through Peer Critique
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What’s Next
ArXiv will begin implementing this policy immediately, with moderators and section chairs reviewing recent submissions for evidence of unchecked AI content. Future submissions will require peer review acceptance, and authors found in violation may face bans and increased scrutiny.
Better Health with AI: Your Roadmap to Results
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
What types of AI-generated issues will lead to a ban?
Incontrovertible evidence such as hallucinated references, unverified meta-comments from AI tools, or blatant errors that indicate the authors did not verify AI outputs can result in a ban.
Can authors appeal a ban?
Yes, authors can appeal ban decisions. The process involves documentation by a moderator and confirmation by the section chair before enforcement.
Will all AI-generated content be banned?
No. Only cases with clear, incontrovertible evidence of unverified AI outputs will trigger sanctions. Responsible use and verification are still encouraged.
How will future submissions be verified?
Future submissions will need to be accepted at reputable peer-reviewed venues before posting on ArXiv, ensuring higher standards for quality and verification.
Does this policy apply to all arXiv categories?
Currently, the policy is focused on the computer science section, but similar measures could be adopted across other categories if needed.
