A leading SciFi/Fantasy fiction magazine is closing submissions because of an avalanche of artificial intelligence-generated spam.
Neil Clarke, the publisher and editor-in-chief of the award-winning Clarkesworld Magazine, said in a Twitter thread Tuesday that he would temporarily close the magazine to new submissions because of the sheer amount of poor quality content written by AI chatbots and other software. Clarke lamented that there are no easy solutions to what could become a growing problem.
“Submissions are currently closed. It shouldn’t be hard to guess why,” Clarke wrote Monday.
Clarke gave more details on the situation Tuesday morning. “We aren’t closing the magazine,” he wrote. “Closing submissions means that we aren’t considering stories from authors at this time. We will reopen, but have not set a date.”
“We don’t have a solution for the problem,” Clarke continued on Twitter. “We have some ideas for minimizing it, but the problem isn’t going away. Detectors are unreliable. Pay-to-submit sacrifices too many legit authors. Print submissions are not viable for us. Various third-party tools for identity confirmation are more expensive than magazines can afford and tend to have regional holes. Adopting them would be the same as banning entire countries. We could easily implement a system that only allowed authors that had previously submitted work to us. That would effectively ban new authors, which is not acceptable. They are an essential part of this ecosystem and our future.”
Clarke identified most of the problems stemmed from outside writers. “The people causing the problem are from outside the SF/F community,” he said. “Largely driven in by ‘side hustle’ experts making claims of easy money with ChatGPT. They are driving this and deserve some of the disdain shown to the AI developers. Our guidelines already state that we don’t want ‘AI’ written or assisted works. They don’t care. A checkbox on a form won’t stop them. They just lie.”
In a separate blog post on his website, Clarke pointed out that the number of plagiarized submissions began to tick up at the end of 2022, and have surged dramatically in 2023 after AI chatbots became more popular. A graph on the post showed that less than 25 submissions per month were rejected and banned for plagiarism in October; that number increased to 50 by December, then surged all the way to 350 by February 15, 2023. An updated graph shows that the number has spiked to more than 500 as of February 20.
Clarke said that these AI submissions have obvious patterns, but refused to give examples because he did not want to help the violators get better at avoiding detection or paint legitimate authors with a broad brush. He also admitted that several stories that were rejected probably evaded AI detectors, or resulted from editors erring on the side of caution. Clarke also noted that other editors of fiction platforms have experienced similar problems; the phenomenon appears to be mostly aimed at high-profile markets that offer more money per word.
Plagiarism in the arts follows multiple instances of plagiarism in education. Last week, a Florida high school said that students in a prestigious academic program were cheating on their essays using ChatGPT. In December, a Furman University philosophy professor said that AI chatbots are the future of plagiarism.