The well - known preprint platform arXiv is set to revamp its review mechanism for computer science submissions. This move comes as a response to the overwhelming "deluge" of review and position papers, a significant portion of which are churned out by AI systems.

Under the new set of rules, articles falling into this category must first gain acceptance from academic journals or conferences. They are also required to go through a complete peer - review process, and authors need to submit solid proof of acceptance. Failure to meet these requirements will likely result in the rejection of the paper.

The surge in the number of submissions can be partly traced back to the widespread popularity of generative AI and large - scale models. The new policy is designed with the intention of easing the workload on reviewers. By relying on external peer - review processes to pre - screen content, it aims to streamline the review process.

It's important to note that this is not an entirely novel regulation. Moreover, it has no bearing on the submission of regular research papers. However, if other disciplines also witness a similar surge in submissions, arXiv may consider extending these review measures to those areas as well.