top of page
Emerge_typography - white.png

AI-Generated Comic Art: Do WeHave a Right to Know?

Aug 16
5 min read

AI-generated comic art is becoming a thing, and there's a lot of heated discussion about it. The big question though is, should creators let us as readers know it's AI-generated? The short answer? Yes. The long answer? As AI transparency laws, platform rules, and community expectations evolve, clear disclosure is becoming both an ethical standard and a practical risk-management move for comic creators.


AI has thoroughly disrupted the creative world

Maybe that’s my own bias talking - as both a creative and a marketer - but it feels like AI is hitting the creative community harder than almost anyone else.

Naturally, it’s crept into comics.

People are using tools like Midjourney to generate pages one panel at a time. More often than not, these are writers sitting on great scripts who lack the small fortune required to hire an artist, colorist, letterer, and editor.

And look, I get it. Money is hard to come by, making comics is brutally expensive, and most indie projects are funded by optimism and couch coins.

Still, the fallout has split the community cleanly in two:

  • The Purists: AI art has no place in comics. Full stop.

  • The Pragmatists: I have a story to tell, no money, and a prompt box. Watch me.

Wherever you land on that spectrum, the comic world at large is starting to draw lines.

Creators, publishers, and conventions are no longer treating AI-generated art as a fringe issue. New York Comic Con, for instance, banned AI-generated products from being made available for purchase.

And now, the lawmakers have entered the chat.

The Law Is Catching Up (Sort Of)

AI transparency rules are starting to take shape, with both the EU and California rolling out laws designed to make AI-generated content easier to identify.

The EU AI Act includes obligations around marking and disclosing certain AI-generated or manipulated content, while California’s AI Transparency Act, SB 942, will require large generative AI providers to support detection tools and disclosure mechanisms for AI-generated image, video, and audio content beginning in August 2026.

The two laws approach the transparency problem differently, but the underlying vibe is the same: Stop hiding the hand.

The EU AI Act focuses heavily on synthetic media and deepfakes - anything that might trick a human into believing a real person said or did something they didn't.

California’s Transparency Act takes a slightly different angle, focusing on watermarks and digital paper trails for generated images, video, and audio. It places the burden primarily on the massive tech providers rather than forcing you to label every AI-assisted piece you create.

This means the AI tools themselves will start baking disclosures directly into files - in fact, Anthropic announced that it’s going to be watermarking text and files generated by its Claude models, and you know the others won’t be far behind.

It also means the platforms hosting your work are likely to make you check a box, add a label, hidden metadata, a note in the credits, or all of the above - the direction of travel is obvious: AI-generated work is getting much harder to hide.

What does this actually mean for creators?

This matters especially in comics because AI detection is already unreliable. In a previous blog about AI image generation, I tested detection tools against both hand-drawn and AI-generated comic pages, and the results were inconsistent. Some refined AI pages slipped through. Some human-made work could easily be viewed with suspicion. 

Many artists are having to record themselves working to prove that they’re not using AI. Their work is too perfect? It must be AI. Their work contains a mistake? It must be AI. The amount of work that artists now need to do to prove that they actually created the piece is growing fast.

Having some sort of watermarking on AI content is going to let everyone breathe easier.

The Hot Potato Strategy

Right now, uploading an AI comic to a platform usually involves checking a tiny box that AI assisted in some way. It feels more like a formality than anything else.

It won’t for long.

As legal fines and reputational risks scale up, platforms will do what platforms do best: protect themselves and pass the liability downward.

If you upload an undisclosed AI comic, the host platform isn't going to take the bullet for you. They’ll point directly at you and say, "They lied to us on form 3B. Blame them."

Transparency isn't just a moral debate anymore. It’s basic risk management.

How Likely Is This to Become a Real Issue?

I’d like to say it’s unlikely.

But honestly? I’ve seen the Reddit threads. I’ve seen the Facebook posts. I’ve seen what happens when someone shares AI-created comic art without being clear about how it was made.

The reaction isn't a gentle debate on authorship. It’s an immediate, enthusiastic mob with digital pitchforks. People get furious. They feel AI art steals from artists, devalues the grueling years it takes to learn visual storytelling, or that somebody’s job was taken from them.

Whether you agree with those arguments or not, the hostility is real.

So if you’re using AI to make a comic, don't be vague. Don't call it "digitally enhanced" when you mean "prompt-generated." Don't try to hide the ball.

Tell people what the machine did.

Tell people what you did.

How Comic Creators Should Disclose AI Use

If you use AI-generated or AI-assisted art in a comic, the safest approach is to be specific. A vague label like “digitally enhanced” creates more suspicion than trust.

A useful AI disclosure should explain:

  • which parts of the comic used AI

  • which tools were used, if relevant

  • whether the images were prompt-generated, edited, painted over, or composited

  • what human creative work was involved

  • whether a human artist, editor, letterer, or designer contributed

“Interior artwork was generated using Midjourney and manually edited by the creator. Script, dialogue, lettering, layout decisions, and final page assembly were completed by the author.”

And then you have to let the reader decide if they care.

Because ultimately, it's their choice. Not the angry commenters', not the platforms'.

The reader's.

Honesty About the Hand

As much as we might want to keep comics pure, it’s not really up to us.

The genie isn’t going back in the bottle. AI-generated work is going to be shared. Some of it will be lazy. But some of it may be interesting - ideas that would never have seen the light of day because there were no coins under the couch cushion.

But that’s for the reader to decide. Not whether it should be disclosed that it’s AI-generated - that part should be obvious. The reader gets to decide whether AI-generated comics are worth their time, their money, and their attention.

And it’s up to us to trust them with that choice.


Sources



 
 
 

Comments


bottom of page