The AI-generated Elephant in the Room
In a Chill Subs survey of 132 editors and staff, six respondents described practices ranging from flagging writers suspected of using AI to banning them from submitting again. One said suspicion alone was enough to place a writer on a watchlist.
Right now, many literary magazines are deciding what’s human-authored largely on instinct, and a writer can be flagged based on a suspicion they may never know about. The publishing industry’s response to AI can’t just be about how effectively it identifies AI-generated writing. It has to be about how fairly it treats the humans whose work is being evaluated.
I think a lot of writers feel the way I do—we’re tired. Tired of talking about AI, tired of talking TO AI, tired of losing sleep (and jobs!) over AI. I’m mostly tired of not knowing if what I am reading/watching/consuming is real, true, or human-created (these are three different things). I’m tired of my attention being depleted by questions of truth and authenticity. I’m tired of constantly being in suspicion mode (did this writer use AI? do they suspect I used AI?). I want to be in curiosity mode, where I get to just engage with ideas (remember those days?).
I’m not going to spend a lot of time on the recent AI scandals in publishing. Instead, I want to pull back the curtain a bit. How is the publishing industry determining what is written by AI? What are the costs of an ever-increasing game of gotcha, whodunit, and demands for proof?

The full reporting and analysis on The State of Publishing Today as it connects to AI sits below the paywall.
Alongside this column, paid subscribers also get weekly access to Jane Friedman on the business of writing, Erin Weiss on which workshops and programs are worth your money, and access to year-round free submissions, plus hours of craft workshops right here on Substack.





