AI Use Policy

Last updated: May 13, 2026

Let’s be clear

Humans Write exists to celebrate human writers writing short form literature. Our mission is to provide an accessible literary platform that thrives through the collective contribution of its community, enabling writers to bravely share their unique narratives, connect globally, and pursue financial rewards for their talents.

In order to make that into a reality, we all need to agree on a few things. If you’re going to participate in this community, you need to understand how we delineate between human written works and AI assisted or prompted works. 

Literature and writing is undergoing a historically significant transformation. One that other industries have quietly undergone slowly, and then all at once were transformed over night. Remember when every singer and vocalist had to have a good voice? And now there’s autotune? People understand what you mean when a musician does an “acoustic” set. And the same people who perform beautifully without assistance often use modern tools when producing their latest album. It’s a mixed world out there. And the literary world is just getting into the mix. 

We, humans, however, are at a disadvantage when it comes to understanding what is an “acoustic” piece of literature versus a heavily mixed, electronically influenced piece. 

With this in mind, we have a policy in place outlined below that helps us and you understand if and when a human or an AI contributed a piece.

The three categories

Whenever you enter a competition or submit work to us, you'll be asked to declare your piece “Human Written”, “Robot Assisted” or “Robot Written”. The same goes for much of the content you see on our site. We do our best to tag and inform readers what fits into these categories. 

Human Written

You, a real human, wrote it. You used your unique voice, your lived experiences, your amalgamation of digested literature read throughout your life thus far. Some clarifications:

• Spellcheck and grammar checkers (including AI-based grammar tools like Grammarly - fair warning though, don’t let Grammarly make all the decisions for you. It’s crafting things in a specific tone that might contradict yours.)

• A human reader who gave you feedback

• Reference materials, dictionaries, thesauruses

• Research tools (including AI tools used purely to find or summarize background information, not to write your piece)

• Note: If you use AI for research, you are responsible for ensuring no AI-generated phrasing from the research tool makes its way into your final draft and no hallucinated facts made it into your submission.

To make this claim, you may not use AI to generate, rewrite, or substantially compose any part of the writing itself. If an AI wrote an entire sentence, that sentence is not considered human written. Please mark your work as Robot Assisted in that scenario. 

Please note: Minor stylistic or grammatical rephrasing by an AI tool (like Grammarly) does not disqualify a piece from being Human Written, provided the core imagery, ideas, and wording remain yours. 

Robot Assisted

You and a robot teamed up. You brought the idea, the direction, the editing eye, and a minimum of 65% of the words (nearly two thirds). If you’re entering a 300 word micro competition, this means you have written at least 200 of the words yourself.  

The AI may have helped you generate a few sentences, restructure paragraphs, or rephrase passages. Sometimes folks need help figuring out a closing paragraph or how to add more mystery to a plot. Short, literary fiction can be hard to master. A little input from a robot pal is ok. 

In full disclosure, this AI Use Policy is a Robot Assisted written work. We’ve written about 85% of it, and put it through AI to ensure it was formatted ok. We took other writings of ours on this topic and put those into a chat alongside this document to confirm we weren’t contradicting ourselves in other places. It’s a useful tool. But we’d be lying if we said it was 100% human written. We also had real humans review it too. A large portion of submissions will fall into this category. It’s an uncomfortable truth that we all must come to terms with. 

Robot Written

You typed a prompt. The AI gave you a piece. You used it as-is, or with only tiny edits, fixing a typo, swapping a word, changing the title.

Like a word calculator: input goes in, output comes out, and you handed it in.

If you put significant revisions in and moved sentences around but did not write any more than a paragraph or two, you are still in Robot Written territory. If you’re working on a 700 word microfiction piece, and only 50% of it is written by you - you are still in “Robot Written” territory. You need to write more than the robot to move into the “Robot Assisted” category.

What you have to do

When you submit a piece to a Humans Write competition or social media platform, you tell us which category it falls into. That's it. We don't penalize Robot Assisted or Robot Written entries, but depending on the competition official rules, Robot Written might be judged differently. And if it is discovered that you lied about your entry’s classification, you will be disqualified from the competition out of respect for those who are participating fairly and honestly. And it is possible you may be completely banned from participating in our platform ever again. 

Making a false declaration, claiming Human Written when it isn't, is the one thing we treat seriously. We don't do this to be punitive. We do it because the whole point of this community is honest writing in a world full of people posing as if they “wrote” something when they simply prompted a word calculator.

And there’s another layer to it. Laws are ambiguous around this right now. In the future, things are going to change. It’s best to be straightforward now, to avoid future issues. Currently, the law generally only protects works with significant human authorship; your honesty helps protect the legal integrity of your own work.

What we do with your declaration

Three things:

1. We use it to judge fairly. Your time spent toiling over the right word to use to describe your main character’s torment is time well spent. We value this. 

2. We publish it.  Robot Assisted and Robot Written entries that get published are labeled as such, alongside Human Written ones.

3. We hope to build something that celebrates human efforts. Humans Write plans to use submitted work (with your consent, as described in the relevant competition's Official Rules and our Privacy Policy) to train a machine-learning model designed to detect human generated text. Not to make derivative works with it. Never to generate it. The goal is to give writers, readers, and editors a tool to help recognize human writing in a world flooded with the other kind. This is a goal at the moment and will take thousands of submissions to get there. But we’re being up front about it so you can be aware of how we plan to collectively celebrate human written works. Currently, models are based around detecting AI written text or mimicking human written works. We’re putting humans front and center to help our culture never lose sight of the human art of writing.

4. No feeding third-party LLMs. While we use your data to celebrate and protect human writing, we strictly prohibit third-party AI companies from scraping this site to train their generative models.

Our own AI use

Because we expect honesty from you, here's ours.

• Customer-service replies are written by humans (you can reply to any of our emails and a real person will write back).

• We do not use AI to write competition feedback to entrants. That's a human-judge job, always.

• We use AI tools for things like spellcheck, occasional image-generation for blog headers, and behind-the-scenes operations. We label public-facing output where the work itself was AI-touched. Posts on our website and social media are tagged: Human Written, Robot Assisted, or Robot Written. Look for the label.

A note on the future

The line between human and machine writing will keep moving. We'll keep updating this policy as the tools change and as we learn more about what entrants find honest, useful, and fair. 

Questions

Email us at hello@humanswrite.com. A human will write back.