A Note from Your Editor Regarding the Entirely Routine Authentication of Your Humanity

Dear Author,

Thank you for submitting your manuscript. Before we proceed, we require confirmation that it was written by you, a person, rather than a technology trained on books written by you without your knowledge or consent.

We appreciate that this may feel circular. Please be assured it is not circular. It is an ouroboros. A spiral of accountability. An ethical feedback noose.

As you know, several prominent authors of color have faced allegations that their novels were written with the assistance of generative AI. Some have lost publishing contracts and any remnant of faith that “innocent until proven guilty” is a part of the acquisitions process.

All have denied using AI. Naturally, these denials cannot be accepted without evidence, since denial is exactly what a guilty person would do, as well as an innocent person, making the two functionally indistinguishable.

Fortunately for us all, we have AI-detection software.

To establish that AI did not write your book, we will upload it to an AI detector. The detector will retain your words, learn from your distinct syntax and semicolon placement, and use them to improve the technology that will eventually produce more books resembling yours.

Results are extremely reliable, provided “reliable” is understood in its newer, more innovative sense. For instance, your manuscript received a score of 79 percent human when pasted into the detector, 33 percent human when uploaded as a Word document, and 7 percent human when uploaded as a PDF. (We believe this reflects the PDF’s suspiciously robotic commitment to preserving formatting.)

Several passages were flagged because they were “too predictable” and/or included “hallmarks” of AI. Others were flagged because they were “too original.” Your use of semicolons was deemed artificial; a desperate attempt to flee from the cesspit of discarded em dashes.

The sentence “She returned home,” in chapter 12 was identified as 92.23 percent machine-generated, presumably because no human being has ever returned home.

We invite you to demonstrate your authorship by providing browser histories going back sixteen months, childhood diaries, school reports, and testimony from the OB/GYN present at your birth. If you wrote early drafts by hand, please submit the original notebooks for carbon dating.

We recognize that some white authors have openly discussed their use of AI. We applaud them for their honesty. This is not inconsistent behavior on our part. Indeed, their candor demonstrates a bold willingness to interrogate the human-machine interface and the possibilities of collaboration and form. Your alleged use of AI demonstrates fraud.

The distinction remains that they admitted to using AI, whereas you deny it. Had you used AI and admitted to it, we might have praised your candor. Since you did not use AI, our options are regrettably limited.

Please understand that this process is not punitive. It is our commitment to protecting literature from authors whose work was used to build the machines used to determine whether those authors wrote their work.

Should your contract(s) be terminated, all rights shall revert to you, except, of course, whatever rights have already been absorbed by large language models.

Thank you for your cooperation as we navigate this unprecedented moment.

To help us reach a prompt decision, please reply by Friday and prove that you are real.

Warmly,
Your Editor

This message was drafted with the assistance of AI.

Please help support our writers and keep our site ad-free by becoming a patron.