ArXiv, the open-access repository hosting over 2 million preprints in physics, mathematics, computer science, and other fields, is enforcing stricter penalties against authors who misuse artificial intelligence in their submissions.
The platform now implements a one-year ban for authors found to have allowed AI systems to generate entire papers or substantial portions of their work without proper human oversight or disclosure. This escalates previous warnings and temporary suspensions, signaling ArXiv's commitment to maintaining research integrity as large language models become increasingly sophisticated.
The crackdown targets a specific problem. Researchers increasingly submit papers where LLMs like GPT-4 or Claude generate core content, methodology, or analysis without the authors meaningfully reviewing, validating, or editing the output. This creates papers with plausible-sounding but potentially fabricated citations, flawed logic buried in dense text, and results that haven't undergone genuine intellectual scrutiny.
ArXiv's moderation team reviews flagged submissions and can now impose year-long posting bans on repeat offenders or egregious cases. Authors caught using AI carelessly lose the ability to submit new preprints to the repository. Given ArXiv's role as the primary distribution channel for physics and computer science research, this penalty carries real consequences.
The policy distinguishes between responsible AI use and lazy automation. Using AI as a research tool—for literature review assistance, code generation, or draft polishing—remains acceptable if authors disclose it and maintain intellectual responsibility. The ban targets wholesale delegation to AI systems where humans abdicate critical thinking.
This move reflects growing tension in academia. Publishers and repositories must balance accessibility of AI tools with preserving peer review standards and research quality. ArXiv's approach sends a clear message to the research community: AI is a tool, not a replacement for the author's brain. The year-long ban creates real incentive for authors to