tech
May 17, 2026
ArXiv introduces one-year ban for researchers who submit papers with unchecked AI-generated content
ArXiv will ban researchers for one year if they submit papers with obvious signs of unchecked AI generation, such as hallucinated references or leftover chatbot instructions. The policy, announced by computer science section chair Thomas Dietterich, is the first formal penalty by a major preprint platform for AI-generated slop.
TL;DR
- ArXiv will ban authors for one year for submitting papers with incontrovertible evidence of unchecked AI generation.
- The penalty is triggered by clear errors like hallucinated references, meta-comments from chatbots, or placeholder data left in the text.
- This policy does not prohibit the use of AI for drafting or editing but targets careless submission of unvetted AI output.
- A study in The Lancet showed a twelvefold increase in fabricated citations in biomedical papers since 2023, attributed to AI writing tools.
- ArXiv's volunteer moderation system was not designed to screen for machine-generated content at scale.
- The policy focuses on obvious errors that prove an author did not read their submitted paper.
- Other academic institutions and conferences are also grappling with a surge in AI-generated submissions with minimal human oversight.
- ArXiv is becoming an independent nonprofit, which may provide greater autonomy for policy changes and fundraising for quality control.
- A new requirement for first-time submitters to obtain an endorsement aims to reduce submissions from accounts created solely for publishing AI-generated material.
- The policy establishes that authors are fully responsible for their content, regardless of how it was generated.