ArXiv will ban researchers who upload papers full of AI slop

arXiv will suspend authors for one year if their papers show clear signs of unchecked AI-generated content, such as fake citations or leftover prompts.

The policy targets low-quality “AI slop” and LLM hallucinations to protect the reliability of preprint research, especially in computer science.

Authors can appeal bans, and future submissions may be restricted to work already accepted by reputable peer-reviewed venues.

The move marks a tougher enforcement stance against misuse of generative AI in academic publishing.