This is a long time coming. I've been posting about the decline of arXiv's CS category for a long time now, and even had a few conversations with someone I know who works there about it. Personally, I think the slop started in 2018--prior to generative AI slop--when the CS category at arXiv began the unsustainable exponential growth in submissions that has continued till today. An increasing number of what amounted to corporate whitepapers and other marketing materials were being posted on arXiv to give them the appearance of scientific credibility. There was a fairly clear arXiv-to-Nature pipeline. Citation counts were pumped as some of the scientometric services count arXiv "articles" as citations, and some researchers adopted the bad scholarly habit of citing arXiv preprints instead of the final publication. It was and still is a mess. My understanding is that arXiv was meant as a place for people to put high-quality but pre-publication articles, but at least in the CS category it's drifted quite far from that.

I gather they've finally taken this measure because of the preponderance AI-generated slop, but with any luck these other issues will improve too. The arXiv press release states “Review/survey articles or position papers submitted to arXiv without this documentation will be likely to be rejected and not appear on arXiv” so it does sound like they are acknowledging the other problems and intend to enforce their rules more strictly in the future.

"arXiv says it will no longer accept Computer Science papers that are still under review due to the wave of AI-generated ones it has received."
From https://infosec.exchange/users/josephcox/statuses/115486903712973154

#arXiv #preprint #CS #spam #AISlop #AI #GenAI #GenerativeAI