Research repository ArXiv will ban authors for a year if they let AI do all the work

ArXiv, a widely used open repository for preprint research, is increasing efforts to address the careless use of large language models in scientific papers. The site has already implemented measures such as requiring first-time posters to secure an endorsement from an established author. After more than 20 years hosted by Cornell, the organization is becoming an independent nonprofit to better address issues like AI-generated content. Thomas Dietterich, the chair of arXiv's computer science section, posted that a submission with incontrovertible evidence that authors did not check the results of LLM generation means the paper cannot be trusted. Incontrovertible evidence could include hallucinated references and comments to or from the LLM. Authors in such cases will face a one-year ban from arXiv followed by the requirement that subsequent submissions must first be accepted by a reputable peer-reviewed venue. The measure is not an outright prohibition on using LLMs. Authors must take full responsibility for the content irrespective of how it is generated, including any inappropriate language, plagiarized content, errors, or misleading information copied from an LLM. Dietterich told 404 Media that the policy represents a one-strike rule, although moderators must flag the issue and section chairs must confirm the evidence before imposing a penalty, and authors will be able to appeal the decision, while recent peer-reviewed research has found that fabricated citations are on the rise in biomedical research, likely due to LLMs.