Café de Nuit

The Fraternity of Shadows’ Position on AI Art

For obvious reasons, we strongly prefer human-created art. The exciting world of D&D relies on the work, imagination, and creativity of people, and this applies to written words as much as it does to visual art. It's not a machine that created iconic D&D monsters such as the beholder, the mind flayer, or famous and complex NPCs such as Strahd.

However, at the level of a fan community like ours, where we publish netbooks for free, we understand that not everyone has artistic skills, knows an artist, or has the budget to commission artwork to accompany an article. In that context, we see AI not as art, but as a « visualization tool », a way for an author to put into images what they have in mind.

Within their own articles, authors are free to choose how they illustrate their contributions. Our netbook values are imagination, ideas, lore, and illustrations can serve all of these purposes.

Therefore, if it is the author’s choice for their own article, we will accept a limited amount of AI-generated art.

Transparency is required

We require complete transparency. Any image that is 100% AI-generated or substantially AI-assisted must be clearly identified as such.

We will not accept anyone presenting AI-generated image as entirely human-created. When in doubt, identify it as AI-generated.

From now on, authors must identify the source of every piece of art used in their article:

- Human-created artwork: provide the artist’s name and source (ex. web site).
- Existing artwork: identify it as a public-domain image, stock image, from deviantart or other borrowed artwork, etc., and provide a link to the original source when it comes from another website.
- AI-generated artwork: identify it as AI-generated (or substancially AI generated if AI was used significantly in the human creative process).

For AI-generated artwork, authors must also provide the prompt used and the software or AI tool used to create the image. For example: “ChatGPT — Make an image of a witch in the forest of Tepest, surrounded by a few fearful goblins.”

No imitation of identifiable artists

We will not accept AI-generated artwork created by directly imitating the style of an identifiable contemporary artist. For example, prompts such as: “draw this character in the style of (Artist ABC)” are not accepted.

You may instead use broader artistic descriptions, such as: “draw in the style of dark gothic fantasy oil painting.”

The distinction is that we are comfortable with describing an artistic genre, medium, period, or general aesthetic, but not with asking AI to directly imitate a specific contemporary artist.

Quality is a must

Sometimes, AI produces weird things like someone with three legs, or similar highly visible flaw. We will not accept AI-generated artwork with similar headache-inducing flaws.

AI is not the product

We also do not want the Fraternity of Shadows website to become a repository for mass-produced AI content.

Please do not submit galleries of AI-generated artwork. We will not accept them.

AI is not the product. AI-generated artwork can be part of an article, in limited number, but it cannot be the article itself.

For this reason, the FoS Gallery will remain AI-free and will feature human-created artwork only.

Identifying AI artwork in our netbooks

In our future netbooks, we will place a small logo beside all AI-generated images to make it immediately identifiable to readers. (to be discussed - a small computer icon, the letters AI, or something similar?).

This will allow readers who prefer not to view AI-generated artwork to identify it quickly and simply skip it.

No AI made written work in our netbooks

Similarly, while we are at it, we will not accept in our netbooks AI written texts, stories, etc.

Respect another’s intellectual property

QtR articles are the sole properties of their authors, and that property is sacred. If you are creating art (AI or not) to accompany an article based of another contributor's article, please ask for the original author’s permission!

These guidelines may evolve

AI technology is evolving extremely quickly. We therefore reserve the right to modify these guidelines in the future as the technology, the community’s opinion about AI, and our understanding of its impact continue to evolve.

At this moment, our goal is to find a reasonable balance: preserving the essential human creativity, while allowing authors to use available tools to bring their ideas, stories, and lore to life.

While I suppose it's a balanced approach, I am concerned that I'll see what has been going on with a different fanzine. Blatant AI imagery has overtaken the magazine for a while, which has discouraged me more and more on submitting. Mind you, I'm not trying to start drama with other zines, so I removed the name instead of posting it here. I'll have to see what comes forth in the next year or so, but I've found myself going from neutral positive to gradually more and more negative towards AI related content.

I share your concern.

As explained above, AI, if chosen by the author, is to be added in limited amount to an article, to help the article's clarity.

Articles with too many un-needed flashy AI will see them removed.

Joël

Any reason to require the prompt for AI image generation? Sound great in theory; it's openness and people can try their own image generations with the same prompt. In practice if you are going through many edits (and this is normal now), the amount of prompt text could fill entire page for a single image. It would seem to penalize image generations that were not one shot and therefore required greater effort.

I am not arguing the matter. I am pointing out that the prompt requirement carries a visual load burden on the page, and does not really add to transparency because the image itself does not carry any particular signature relating to the prompt. If you wanted to stop artists being copied, the best way is the ban certain image generators that allow it. For example, with Google image generators, the prompt is irrelevant. Even if there is an instruction to copy a particular artist, the model will not comply and will produce a generic image instead.

At least we would be able to review and judge the prompt.

We haven't thought about it, but indeed, perhaps the prompt will kept in a separate "source" document, so we can review it as editors, but it won't appear in the netbook.

Another thing: not everything graphical has a prompt at the root. That was the case perhaps even 12 months ago, but now AI can use python to 'hand write ' maps, can use Photoshop equivalents to format graphics, structure it, place and edit via computer use. The request for a prompt is carrying more weight than it should (and fundamentally unverifiable). My understanding about how this policy was formulated is follows:

1. Someone goes to ChatGPT (etc) to and writes a prompt.
2. Image comes out.
3. Done.

This is a basic workflow but not the workflow of an AI developer or experienced person. It's not taking into account the use of AI harnesses, agentic workflows, or work that does not follow a specific prompt.

I understand the need to weed out AI slop. And the need to not publish cloned AI art which is signature to a particular artist. The best way is to just weed out anything that looks low effort. If it looks like someone could have belted it during their lunch break, just exclude it.

But also, at the same time, try it out yourself. See if you can go to ChatGPT with a prompt and produce a similar work right off the bat to test your suspicions of low effort AI slop. If ChatGPT just belts out an equivalent just like that, you've got your slop indicator.

It's all up to the dark masters of the Fraternity, of course. But you may end up in a rabbit hole if the AI policy is misaligned to the actual AI situation. Alternatively, AI could be banned for everything except for clearing up typos and grammar. That would neatly sidestep the problem.

Sometimes, AI produces weird things like someone with three legs, or similar highly visible flaw. We will not accept AI-generated artwork with similar headache-inducing flaws.

I object AI art in general, but I have to admit that a Ravenloft fanzine might be the one place where this sort of 'flaw' could, in fact, serve the art. Like, in a body horror article. Or when a caliban OC is involved. Or, for the non-living equivalent, when someone is trying to create a 'liminal space' scenery that's intentionally 'off' and inhuman.

Well done, friends. This is the way to go with AI.

A compromise I'd prefer, even though I abhor and loath algoslop based imagery, is that the final product not be AI but just a step in the process. If someone needs it for a basis to build art from? Okay, fine. Even just glorified trace over with human input alterations to make it not look like AI. Sure.

So few are willing to do so.

doctornecrotic wrote: Sat Oct 03, 2026 10:50 am A compromise I'd prefer, even though I abhor and loath algoslop based imagery, is that the final product not be AI but just a step in the process. If someone needs it for a basis to build art from? Okay, fine. Even just glorified trace over with human input alterations to make it not look like AI. Sure.

The problem there being that the line between the two is difficult to draw: Just this month, there is one book on DTRPG's top-twenty sales chart that is completely and entirely layouted and has its artwork done with AI. That trend is going to become stronger. -- As to modern layouts being done with AI, that's just run-of-the-mill, industry standard nowadays. Might not be called like it, in each and every instance, but is just that. From Photoshop to GIMP to how your text is positioned against the page. No escape. ...Which makes this a very Ravenloft-y thing if you think about it.

Rock of the Fraternity wrote: Sat Oct 03, 2026 11:46 am So few are willing to do so.

Generative AI as its proponents and creators define it replaces "art" by dispensing with the very process of creating art. Thus, to incorporate AI into the process of creating art defeats its very purpose. The wannabes and the AI Bro bros who vociferously tout AI as the inevitable, all-destroying wave of the future (of art) never do so because incorporating AI into art requires a level of effort and focus and competence they have no ability nor interest in meeting, and very few creatives bother to advertise that they use AI in their works for fear of rightly attracting negative focus.

Baron Von Stanton wrote: Sat Oct 03, 2026 4:29 pm

Rock of the Fraternity wrote: Sat Oct 03, 2026 11:46 am So few are willing to do so.

Generative AI as its proponents and creators define it replaces "art" by dispensing with the very process of creating art. Thus, to incorporate AI into the process of creating art defeats its very purpose. The wannabes and the AI Bro bros who vociferously tout AI as the inevitable, all-destroying wave of the future (of art) never do so because incorporating AI into art requires a level of effort and focus and competence they have no ability nor interest in meeting, and very few creatives bother to advertise that they use AI in their works for fear of rightly attracting negative focus.

We still have not seen the Mozart of AI music, Michaelangelo of AI painting. But we have seen the AI paragons of pure math, which has been like a nuclear bomb going of in that field of study. There is a narrow focus on how AI affects artists, while little thought is given to how it affects other people, and far more severely than the art field. We will likely have huge breakthroughs in technology and material science thanks to AI math, and no one will think to reject these advancements just because they were created by AI.

Nor do I ever expect to see an LLM Mozart or Michelangelo. If a true artificial intelligence is ever created, then perhaps.

In the meanwhile, this thread is not meant to argue the uses of LLMs in the field of maths, or the way it's affecting other fields. It is specifically about the use of LLM-generated imagery and text.