arXiv will now ban authors for a year if they submit low-quality AI-generated papers

arXiv has updated its code of conduct to confront the growing presence of low-quality, AI-generated submissions, commonly described as AI slop. This update, announced by Thomas G. Dietterich, chair of the computer science section, focuses on ensuring all authors take full responsibility for every aspect of their work, regardless of whether content is produced by artificial intelligence tools.

Under the revised policy, if generative AI introduces plagiarized content, biased material, inaccurate references, or includes hallucinated citations and meta-comments from language models, the authors are held accountable. The rule features a strict one-strike system: if clear evidence appears that results from language models have not been thoroughly checked by the authors, those authors will face a one-year ban from submitting to arXiv. After this period, any future submissions must first be accepted by a reputable peer-reviewed venue. While enforcement is strict, decisions can be appealed, offering procedural fairness to accused researchers.

These steps follow major disruption in arXiv’s computer science review operations. In November 2025, the review section was closed due to a flood of AI-generated papers that often failed to provide novel research findings. This surge in AI submissions has also exacerbated an evaluation bottleneck, overwhelming traditional peer-review capacity.