arXiv, a major pre-print platform used by researchers across scientific fields, says it is tightening enforcement against submissions it identifies as “AI slop”—papers that contain errors considered clearly generated by artificial intelligence. According to reports, the platform’s policy changes focus on authorship and responsibility: researchers who place their names on such papers may face penalties even if the work includes AI-generated content that the platform deems unreliable.
The announced consequences include a year-long ban from using the arXiv system for affected researchers, along with additional ongoing restrictions after the ban period. The aim is to reduce low-quality or problematic submissions and protect the usefulness of the pre-print archive for the research community. While the details of how arXiv determines that errors are AI-generated are not fully specified in the coverage, both outlets describe the same core enforcement approach and timeline of restrictions.
Overall, arXiv’s move represents stricter oversight of how AI tools are used in manuscript preparation and submission, emphasizing accountability for the accuracy of published pre-prints.
