Generative AI Policy

published 2026-09-23 | tags: #gen-ai

This is my personal Generative AI Policy. Unless otherwise specified, this policy applies to all content on this website, or the media this was linked alongside.

0. Scope of Policy

This policy promises limits on how much, if any, Generative AI is used in published creative works like blog posts, video games, websites, written fiction, images, videos -- you know... art.

1. Where Generative AI may be used

Generative AI may be used in the following ways in the production of creative works:

1. A tool for researching, reviewing, or architecting code.
2. Generating or refactoring tedious, time consuming, or error prone code.
3. Editing pros for written content like blog posts.

This doesn't guarantee Generative AI will be used in these ways, just that it may be used for these purposes.

Guiding principles:

• I can do these things myself, this just gets me there faster
• While code can be creative product, I am OK with leaving specifics to an LLM

Example: I used an LLM to upgrade my game to the latest version of the Bevy game engine after reviewing the change announcements

2. Where Generative AI is not used

Generative AI may never be used to generate neither final art nor placeholder art; including visual, audio, and written pros.

Guiding principles:

1. If the audience can see it or hear it, a person should be making it
2. If you can't do it yourself, you shouldn't have an AI model do it for you

Example: "It's better to have crappy MS Paint art than to have nice clean AI Generated art"

3. Disclosure

Any time Generative AI is included in the final product in any way, that usage must be disclosed. Disclosure should be specific about what was generated.

Guiding principle: The audience should not need to "guess" if something was made using AI

Example: Generative AI was used to edit the the opening of this book

Not the policy, just some more context

I hate to say it, but LLMs are neat. They have strengths and weaknesses, but I have to admit on the whole they're pretty cool tech.

Having used LLMs a lot for work and for personal projects, I feel well informed about the strengths and weaknesses of LLMs as they stand in late 2026.

I also have eyes and ears and use the internet, so I know how Generative AI can be abused to produce spam, and how annoying it is as a viewer to realize you've been fed Gen AI Slop.

For stuff I'm putting out into the world I think it's important to be up-front about how I may use these tools.

This policy is meant to set some personal boundaries around where AI is allowed, where it's never allowed, and anything not covered by the policy will be covered once I cross that bridge.

That said, this policy has nothing to do with how I may use Gen AI tools for personal/private/unpublished projects. It's important to make space for experimentation, but not at the risk of alienating your audience and jeopardizing your reputation.