We Unknowingly Published AI Slop, and We’re Taking Accountability

Regarding two of the stories we published in March 2026, Issue #26 of Solarpunk Magazine, “The Mycelium Daughter” by Bella Chacha and “The Baobab Data Tree” by Irene W. Collins: Thanks to the work of a number of editors within SFF fiction, a substantial amount of evidence has been gathered that these two authors are 

1) the same person, and 
2) using generative AI to produce their stories. 

There is nothing inherently wrong with an author using multiple pen names. However, this author submitted stories to our magazine under two different names, while withholding the fact that they are the same person behind both pen names. If we had known that, at the very least we would have published the two stories in different issues, rather than the same issue. Regardless, they misrepresented themself and violated our submission policy. That doesn’t sit well.

As publishers, we’re inundated with submissions that are potentially AI generated, and it is becoming more and more difficult to identify a piece of AI writing simply by reading it. As a result, we’re vigilantly on the look out for tools that will help us identify and weed out AI generated work. So we completely understand why publishers have started using AI checkers. In many ways, it feels like there isn’t much other choice.

At the same time, here at Solarpunk Magazine we are very wary of AI checkers. We do not use them, and we believe it is unethical for publishers to use them, especially without permission from authors. AI checkers are also generative AI programs. While they aren’t as energy intensive as LLM text generators, we believe it is a hypocrisy for publishers to use AI checkers (especially without a clear policy stating that publishers may run authors’ work through an AI system and permission from authors to do so) while insisting that authors are not allowed to use AI themselves at any stage of the writing or brainstorming process.

Don’t misunderstand, this is not a suggestion that publishers should start accepting AI generated work. Quite the opposite, we’re only suggesting that AI checker AI is as unethical as LLM generator AI, and that publishers should rely on other methods to determine whether a given work was or was not generated by AI. It is also true that other publishers, so far as we know, don’t rely solely on AI checkers in determining the authorship of a work. Still, we stand opposed to the use of AI checkers at all.

Rather than using AI checkers without author consent or knowledge, we rely on our own knowledge of the writing craft and the conversations and interactions we have with authors to help us determine if a piece of work was created using AI. That’s not a perfect method, but neither are AI checkers, which are known to not only produce false positives, but to produce false positives at higher rates for writers of color, ESL authors, and neurodivergent authors.

Early in the process of assessing the writing, we had made the conscious decision to not take action against the writer; spec-fic publishers who raised initial concerns were transparent about their use of AI checkers as diagnostic tools and, as we’ve noted here, we find those tools unreliable and problematic. However, much more evidence than AI checkers has been gathered since then in regards to stories supposedly written by Bella Chacha, aka, Irene W. Collins, aka Charity Ogechi, aka Victory Fortune, aka Ferdinand Ugochukwu. That evidence includes the refusal to discuss craft with editors or to provide story notes and early drafts that are part of every human’s writing process and that would go a long way to prove these stories were written by a human. When we tried to contact Bella/Irene to have a similar conversation, we were ignored.

We have removed Bella Chacha and Irene W. Collins stories from Issue #26, and we are in the process of resending the issue to all our subscribers, and to anyone who has purchased a copy of Issue #26. 

We want to assure our readers and the authors we publish that we do what we can within ethical reason to weed out AI authorship from our submission pool. While it’s entirely possible we won’t always get it right, and that we’ll make mistakes again in the future, we will always be transparent about and be accountable to our community of readers and writers for our mistakes.

In solidarity,

The Solarpunk Magazine
Editorial Team