E106: AI-Generated vs AI-Assisted Art: What Creators and Customers Need to Know

What happens when artwork you created yourself is suddenly labelled as AI?

In this episode, Kaylie and Delores explore the increasingly complicated difference between AI-generated and AI-assisted creative work. 

They discuss reports of Instagram adding AI information labels to content edited with Canva’s background remover, why a single label can misrepresent hours of genuine creative work, and the effect that misunderstanding can have on an artist’s reputation.

They also have an honest conversation about where they personally draw the line with generative AI. 

Is using an AI-generated image for a temporary marketing graphic different from selling generated artwork as your own? 

Does every use need to be disclosed? 

And how can small creative businesses make practical choices without dismissing the genuine concerns surrounding consent, copyright, creative jobs and trust?

Kaylie and Delores also explore why visibly human design—complete with wobbly lines, handmade marks and imperfect details—may become even more valuable as generated content becomes easier to produce.

In this episode:

• The difference between AI-generated and AI-assisted work
• Why Instagram’s AI labels can mislead customers
• How Canva’s automated tools may trigger AI information labels
• The difference between creative authorship and automated editing
• Using generated imagery for marketing versus products for sale
• The pressure placed on solo creative business owners
• Why context, transparency and kindness matter
• Etsy’s evolving approach to originality, digital products and AI
• Why handmade and visibly human design may become more valuable
• Practical ways artists can share their process without constantly defending themselves

This is not an episode about choosing a side and shouting at everyone on the other one. 

It is an honest, nuanced conversation about creativity, technology, trust and the human beings trying to navigate the messy space between them.

Listen now, share the episode with another artist, maker, author or creative business owner, and let us know where you draw the line between AI assistance and AI generation.

Support the Creative Juggle Joy Podcast

Transcript

Copy Transcript Copied

Kaylie Edwards: [00:00:00] Imagine this. You take your own photograph or you make a piece of artwork yourself. You put it into Canva, remove the background, upload it to Instagram, and Instagram adds an AI info label.

Delores Naskrent: And that person has no idea what the actual meaning of that label is. They probably aren't thinking, "Oh, she must have used a little automated tool to remove the background." They may just simply assume that whole image was generated." 

Kaylie Edwards: Exactly. And a tiny label can suggest a completely different creative process from the one that actually was being used.

Delores Naskrent: And for the artist, that's pretty, a pretty big misunderstanding. Trust is such an important part of what we do. People care that we made the work. They care about how we made it, and they care about that little bit of us that they can see in it. 

Kaylie Edwards: So today we are talking about the difference between [00:01:00] AI-generated and AI-assisted work, the uproar around the labeling that has been happening of recent, and why handmade and visibly human design may be becoming more valuable, and why this whole transition needs a little more accuracy and a little more kindness and compassion.

Welcome to the Creative Juggle Joy Podcast, the show for artists, makers, and creative business owners building something meaningful while juggling real life alongside it. I'm Kaylie Edwards, founder of Spellweaver Creative Studio. 

Delores Naskrent: And I'm Delores Naskrent. I'm an artist, designer, and educator, and I help other artists build their skills, their confidence, and hopefully a creative business that really works for them.

Kaylie Edwards: Yeah, this is not going to be a pick a side and shout at the other one episode. 

Delores Naskrent: No, definitely not. I think we can have really real concerns about generative AI without treating every tool, every use, and every artist's situation as though they are exactly the same. [00:02:00] 

Kaylie Edwards: Yes. We should begin with the language, because someone listening may never have heard the distinction between AI-generated and AI-assisted.

AI-generated usually means the system creates the text, the image, the music, or the video itself from a prompt. You might type, "Create a whimsical poster of an artist painting in a garden," and the tool invents the image. AI-assisted means a person has already created the main thing, and the software helps with part of the process of editing.

That could be removing a background, selecting a subject and moving them, correcting noise and it... or transcribing audio or checking text. 

Delores Naskrent: So maybe the simplest way to look at this is did the software actually make the creative content or did it just help the artist do something with work that they had already created?

Kaylie Edwards: Yeah, that is a useful starting point. It is not perfect because generative fill can invent only one part of a [00:03:00] human-made image. But it immediately shows why one label cannot explain everything. 

Delores Naskrent: And as an artist, I think all of those decisions really matter. If I draw the artwork, if I choose every shape and every color, and then use a tool to lift it off the background, that doesn't suddenly mean that the software created my art.

I still created it by hand. 

Kaylie Edwards: Yes, and whereas if the image generator makes the main illustration and I place text over it, the balance of making is different. 

Delores Naskrent: Now, saying those things are different doesn't actually automatically tell us whether each use of it is right or wrong, but it certainly tells us we need something a lot more accurate than one big label that simply says AI.

Kaylie Edwards: Yeah, maybe the useful categories are fully generated by AI, partly generated, AI-assisted edits, and no generative AI used. [00:04:00] 

Delores Naskrent: I think even those four categories feel clear to an ordinary customer, do you think? Is there a better phrase than AI-assisted for routine automated editing? I think part of the confusion is that when we talk about AI as though it suddenly appeared yesterday, when forms of automation have been hiding inside the programs we use for years 

Kaylie Edwards: Yes, algorithms and machine learning have been underneath everyday software for years.

Google introduced RankBrain, a machine learning system used in search in 2015. Search engines have long used algorithms for ranking, prediction, spam detection, and recommendations. 

Delores Naskrent: Exactly. None of us said, "I used artificial intelligence to find that." We just say, "I Googled it," and we carry on with our day.

Kaylie Edwards: Exactly. Design software also automated difficult jobs long before the current wave of image generators. Photoshop introduced content-aware [00:05:00] fill in 2010. It could remove an object and use the surrounding pixels to fill the space. Automatic selections, masking, retouching, and noise reduction have all become normal parts of creative software.

Delores Naskrent: And the artist still had to decide what to remove, what to keep, whether the result was any good. The tool may have saved us a little bit of the fiddly work, but it didn't necessarily come up with the artwork in the first place, right? 

Kaylie Edwards: Yeah, and that's the big change is in modern generative AI. It can produce a new expressive material from a prompt, and it was trained on enormous collections of existing work.

That brings serious questions about consent, copyright, imitation, and jobs. 

Delores Naskrent: Yeah, and those concerns are absolutely real. When we call removing a background and generating an entire illustration the same thing though, we aren't making the conversation more ethical. We're just simply making it less [00:06:00] accurate 

Kaylie Edwards: Yes.

And we make it, it almost impossible for beginners to know what anybody is objecting to. Which automated features have you both used for years without feeling that the software actually became the creator? 

The reason for this episode is because we've been seeing so much about this of late.

And this came onto my radar recently from a carousel post from a business strategist, Gemma Bonham-Carter. She discussed creators finding that posts involving Canva's background remover were receiving AI information label on Instagram. A lot of people in the comments said that they had seen it too and were feeling it from themselves.

There was authors that were having their work said as AI labeled, and they weren't doing that with AI. They had created it themselves, and they were just using the editing features within Canva. And that is useful first-hand evidence that creators are experiencing a problem right now. It is not proof that every Canva export is [00:07:00] affected.

Delores Naskrent: Yeah, and that consistency is what makes this so scary and frustrating. I have recently recommended to people to use the Canva background remover, and they would take something into Canva, take the background out, and then bring it back into their artwork. If two artists seem to use the same tool and one gets labeled while the other doesn't, how on Earth are we supposed to know what is safe to use?

I'd be scared then that piece of artwork would then be identified as having been AI generated in some way. 

Kaylie Edwards: Yes, and there's been obviously surveys as well, and people have seen that if it has a label on it and it says something about AI, people will scroll straight past it, and the reputation will be tarnished for people that have only just used it editing for their original work.

And KQED also reported in August [00:08:00] 2026, this month, that Instagram marketers had noticed Meta's systems appearing to flag content made with Canva background remover. Meta says it relies on technical signals, including industry standards such as a C2PA, as well as information people disclose themselves. In plain English, files can carry information about how they were made or edited, and platforms can read those signals.

Delores Naskrent: So it isn't even looking at the picture and understanding the way you and I would if we were looking at it, that the artwork was drawn by hand. It may simply be finding something, deep in the metadata of that file that says it was an automated tool that was used or touched it somewhere along the way.

So that is a concern. 

Kaylie Edwards: Yes, and Meta's published approach now distinguishes generated content from content detected as edited with AI. Edited content information may be less prominent inside the [00:09:00] post menu. That is more nuanced than its old made with AI wording, but most customers do not know the technical distinction.

If

they see AI info, they can still assume the whole image was generated. 

Delores Naskrent: Imagine one of my students paints a beautiful floral motif, scans it, cleans it up, and removes the white paper in the background so she can put it onto a colored background, let's say for making a repeat pattern. She may have spent hours painting that flower, but if the finished post gets an AI label, lots of viewers won't see the hours of work that she spent on it at all.

They're just gonna see that label. 

Kaylie Edwards: Yeah, and it's such a concern. There was a photographer that took their original portrait and used an automated tool to tidy the background, and the photograph is still theirs, but the label was flattened, and the whole process into one frightening word that they had this label for AI.

And obviously there was concerns from customers. [00:10:00] 

Delores Naskrent: Yep. We've already seen how misleading that can be. Back in 2024, photographer Pete Souza had a photo from the 1984 NBA finals labeled as made with AI on Instagram. This photograph he had taken decades before, like way before generative AI even existed.

Meta eventually changed the wording after people quite rightly pointed out how misleading it was 

Kaylie Edwards: Yeah, that example helps because it shows that technical provenance and public meaning are not always the same thing. 

Delores Naskrent: And honestly, artists shouldn't have to become experts in metadata just to put a picture of their work on Instagram.

Kaylie Edwards: Yeah, and you've gotta remember as well is most customers aren't savvy with AI. A lot of them can't tell the difference. There's... I'll tell a story. I live in a village. We don't even have a shop. [00:11:00] I live in the mountains of Wales, and our newsletter bulletin is run by volunteers, and the person who's a volunteer is probably a bit more tech savvy, and they used AI to generate the event posters.

But I told the difference, but nobody else could tell the difference. But obviously, that's because I use AI all the time, and I see AI everywhere, and the fact that I could see that in a little village in Wales, and the people in the village couldn't tell that and, the elderly people aren't gonna tell that it's AI-generated the difference between AI-generated and handmade or, designed themselves.

So you've gotta remember is that if they see AI info label on, they're gonna think that is AI. They're not gonna realize the difference between that was finished and then edited with AI. 

Delores Naskrent: Yeah, for sure. There are really three separate things going on here. The label being added, then the platform actually limiting or removing the post because of the label, and then the [00:12:00] audience seeing that label and reacting to it.

We shouldn't assume that those three things are the same. 

Kaylie Edwards: Exactly. The first and the third are much easier to support in this Canva discussion than a universal claim about reduced reach. 

Delores Naskrent: But even if Instagram hasn't formally penalized the post, the artist still pays a price, in my opinion. So customers question the artists in their comments, or they just simply assume the work was generated and they scroll right past.

That can still damage the trust that artist has worked so very hard to build. 

Kaylie Edwards: Yeah, and Canva itself has woven AI throughout the platform now, and Canva says individual users cannot fully switch Canva AI off, although some administrators can control particular features in business plans, which is obviously frustrating and, I can see why there's so much uproar when people don't want to use AI and they can't switch it off in a platform that they have been using [00:13:00] for years.

Delores Naskrent: Yeah. It's incredibly frustrating for artists who are deliberately trying not to use any generative tools. Many of them have used Canva for years to just help, let's say, lay out the post or add a little bit of text. They should be able to see very clearly which tools are generative, turn them off if they want to, and then know that information is not being stored or attached in the metadata of that particular file that they're exporting.

Kaylie Edwards: Yeah, and there's a lot of similar worries among authors as well using Canva for book covers, interiors, and marketing. We did not find evidence that Amazon automatically treats every Canva-made design as AI generated, but Amazon KDP's published rule is actually clearer: disclose AI generated text, images, or translations, and AI assisted content does not need disclosure.

Amazon defines assisted content as work created by the person and then edited, refined, checked, or improved with an AI tool. [00:14:00] So there are some gray areas where people have been banned from KDP or had their books rejected because they've used Canva features and may have not edited it further than they were supposed to.

Delores Naskrent: Yeah. On paper, a policy like that sounds sensible really, but the trouble is whether the automated systems and the people dealing with appeals apply it consistently. So if someone or something is questioned, the creator needs to know exactly what triggered it. 

Kaylie Edwards: Yeah, because AI concern, licensing concern, duplicate template, or not sufficiently original are four different problems, and saying only Canva or AI helps nobody fix the right one.

Delores Naskrent: And this is probably where things get a little uncomfortable, honestly, because now we're talking about somebody being labeled by mistake. Sometimes a creator has knowingly chosen to use a generated image

Kaylie Edwards: [00:15:00] Yeah, let us use a specific scenario. An artist is running a workshop. Instead of spending hours designing a temporary event poster from scratch, they generate a decorative image, add the workshop details, and use it in an email or a social post. They are not selling that image as their artwork. Is that wrong?

Delores Naskrent: Some artists will say that is wrong, full stop. They may be concerned about how the system was trained, the environmental costs, work being taken away maybe from illustrators and designers, or simply that the message it sends when an artist uses generated imagery is not, I don't know, ethical. I don't think any of those concerns should be dismissed.

Kaylie Edwards: No, and the other side is the reality of solo businesses, like myself. The same person is making the art, teaching or filming courses, answering customers, writing newsletters, and doing the marketing. A quick graphic may be the difference between promoting the event today and [00:16:00] never getting it out.

Yeah. And that's where a lot of the compassion needs to kinda come in, is when you're solo in business, you don't have the time to go into Canva for hours or days creating all the marketing graphics yourself and be able to promote it. Most people never get it out, like myself 

Delores Naskrent: Yeah, most of us running little creative businesses don't have a designer sitting in the next office waiting for that assignment.

I do most of it myself personally, but, people have the attitude just hire somebody," and it sounds so simple, except for the fact that, hey, you gotta pay for that. The... Maybe you have a deadline too. Maybe you just don't have the time to get something to a graphic designer, wait for it to come back, and how much time it can take just to find that person to actually take on the work and do it the way you want that work to be done for your brand.

Kaylie Edwards: Yeah, because, like myself, like I have a specific [00:17:00] way I like to make my graphics, my branded graphics. And even sometimes with AI, like I go to create it and it doesn't create the way I want it to look. There's only one that actually creates it the really, the way I want it to do, and it doesn't look as AI-generated as well as some others do.

And it would take me hours in Canva, and I am a perfectionist, unfortunately. And when I create, I like... I have a graphic design diploma, and I go in and I will take ages to do something, and eventually I will just, I'll go, "Sod it, I'm not doing it," because it's taking too long. Or, it could take me hours to create a carousel post if I'm left to it on my own without a deadline.

Yeah. And, the fact that AI can generate it for me, I know I'm not selling it, I'm just using it to promote an event, then I will do that because it's just, it's... I haven't got the money to hire somebody to do it, and I haven't got the time to sit there for hours doing it myself. 

Delores Naskrent: Yeah.

Kaylie Edwards: And this is something we experienced how intense this [00:18:00] disagreement can become for people. A subscriber once sent Delores a long, very critical email because an AI-generated promotional image appeared in her newsletter for the Working Artist Studio session. 

Delores Naskrent: Yes, I remember that one. And that, in that particular case, the guest artist had created the promotional graphic for her own session.

We weren't presenting it as my artwork. It wasn't her portfolio piece, and it wasn't what she was coming in to teach. It was simply an event graphic we had that we were given to tell people about the session. 

Kaylie Edwards: Yeah, and as far as we remember, this was one person raising it, not a flood of complaints.

But the email judged much more than one marketing graphic. It questioned Delores's integrity as an artist, which I was very annoyed about. Yes. But I didn't reply to her. 

Delores Naskrent: Yes. I have to be honest, that email really stung. It really hurt. It wasn't somebody simply saying, [00:19:00] "Delores, I'm uncomfortable with that image.

Could you tell me why you chose to use it?" It was mean. I could have understood that, I think, if she had asked nicely, and I would've been willing to talk about it, but instead it felt as though this one image had been used to make a judgment about my character and my integrity as an artist. I've been an artist and designer for more than 40 years, and I've spent most of my life teaching and encouraging other artists.

I don't take concerns about artists' rights lightly. But I'm also running a very small business and it's a very busy business, and that particular image was just a short-lived graphic that went on one email. It was supplied for a guest session. It wasn't artwork I was claiming to have made. What I wish she had done was to ask me about it before assuming the worst.

A simple question would've opened the door to a thoughtful conversation. The tone of [00:20:00] that email made me feel as though there was no room for context, no room for mistake, and really no room for me to be a human being. It was nasty. We can disagree with somebody's choice without tearing down the whole person 

Kaylie Edwards: Yeah.

And if a large company can easily afford human illustrator but replaces that work with a generator, listeners may judge that differently from a one-person business making a short-lived social graphic. 

Delores Naskrent: Yes, absolutely. And at the same time, we have to be honest about the context. If a generated image could easily be mistaken for the artist's own work, then that needs to be made clear, of course.

And if the generated image is the actual product being sold, that's very different in, meaning and in conversation from using it as a temporary piece of decoration

Kaylie Edwards: okay. Questions to explore together. Is generated imagery more acceptable for a [00:21:00] temporary marketing asset than for the paid creative product? 

Delores Naskrent: In my opinion, yes. It definitely makes my job easier. There's a lot of reasons for me to use generated imagery in that kind of a situation. 

Kaylie Edwards: Yeah I agree there.

Delores Naskrent: Yeah. I don't think I would create something to sell or pass off as my own with, generated imagery personally. 

Kaylie Edwards: Yeah. I think part of it as well is like we buy assets from people, from other artists, and we use them in our own work sometimes. So it's like where do you draw the line?

Delores Naskrent: Absolutely. 

Yeah. And I remember back in the day when I was first a graphic designer, we would have something called clip art. So if you've ever wondered why it's called clip art, it's because we had literally giant [00:22:00] catalogs of stuff that we could clip out with a pair of scissors and paste up.

That's where the term paste up came out in the first place. And you would put together an ad based on some of these things that you had bought, and that's exactly what's happening in this day and age. People don't think of it the same way, so they might consider the fact that you as an artist, you're using let's say a flowered brush set that you purchased from another artist, that somehow it's different than what it used to be when we used to clip it out of a catalog.

I personally produce usable assets for other artists, and part of what I teach and suggest to all of the artists that use my clip art-type products or stamps is that they change it, make it their own, do something with it that makes it original to them. 

Kaylie Edwards: [00:23:00] Yeah. I agree. Should it be disclosed when it is decorative rather than the work being sold?

Delores Naskrent: Personally I'm thinking, "Why would I need to?" I'm also at the same time using a font that is the bulk of the text of an email that was not created by me. Do I need to at the very beginning of the email say, "By the way, I didn't produce this typeface. I bought this typeface that I'm using"? Where do you draw the line, right?

Kaylie Edwards: Yeah. I agree. We use bought fonts from other artists, from other calligraphers in our work. We have built-in fonts from obviously the software, they have built-in fonts that you can use. It's like should we hand draw everything ourselves?

Should we hand write everything ourselves for it to be classed as art? I think there is a very shady gray area with what can be critiqued or criticized. 

Delores Naskrent: [00:24:00] Exactly, yeah. 

Kaylie Edwards: Would a photograph of the real artist, studio, or work in progress be a faster human alternative?

Delores Naskrent: Possibly. But a lot of times, let's say I'm producing something to use within my newsletter I'm not trying to pass it off as my own artwork, but it's the fact that it has decorated or made my email more interesting. I think that's just part of it. I really don't see the distinction of having to identify that.

I think it's just there. 

Kaylie Edwards: Yeah. It's like I don't have time to take a load of photographs for my newsletter all the time to then put in to make sure it's visually appealing and relevant as well to the content that I'm actually creating. Some of it's not relevant to what I'm actually talking about.

So it's easier for AI to generate for me a branded graphic that would work [00:25:00] for my newsletter that's around the topic itself, like instead of me trying to figure out how to photograph or create an image from scratch or use stock photography 'cause I can never find anything that's branded or on point for the audience that I'm actually selling to.

Delores Naskrent: Yeah. 

Kaylie Edwards: And when does business needs speed explain the choice, and when does it become an excuse for- 

Delores Naskrent: I think if you are creating something that you're selling as your own, let's say for example, a greeting card. If I am the artist creating a greeting card, but then I get ChatGPT or whatever software that I'm using, AI, to create the flower that I'm going to put on the card, then in my opinion, it's not really my art anymore.

If I have created the flower, I've painted [00:26:00] the flower, and then I ask ChatGPT to change that for me or add something, maybe do a different color, then I don't know if that's really wrong, honestly. And I really have a hard time discussing this with people because people are very on one side of the fence or the other.

Yeah. Many will argue absolutely not, you cannot ever do that. I personally don't because the fun of it for me is the drawing or the painting or, the layout that I'm creating. But then I'm also creating assets and things that I sell to artists so that they can make those things. So I can't really be on both sides of the fence on this one.

It's pretty hard to, I don't know, to just morally be saying to [00:27:00] somebody who's buying my stuff, "But it's not really your stuff." Yeah. Yes, it is your stuff. If you used my brushes and you created a gorgeous layout and a card that you've put all of your layout experience and your tastes into creating, then it's yours.

It's your artwork. I may have created that original stamp, but that doesn't mean you are not creating the artwork. Yeah. I don't know. Does that make sense? 

Kaylie Edwards: Yeah, no, it does make sense. I think obviously with the speed of business, like you can't be slow with business if you wanna be successful, and there is points where it does annoy me where I see businesses built on generated AI clip art, and they're just selling generated AI clip art collections.

And it does annoy me, 'cause then I think, yes, it would be very quick. Yes, it could be very successful. And yes, I could do with the money. But [00:28:00] I don't want to go that route because I feel that is a disservice to artists. It's obviously a disservice to myself 'cause I feel really guilty for doing something like that.

Delores Naskrent: Me too. 

Kaylie Edwards: And I do see it, and I see ... I follow some businesses as well, and they've created that income stream, but their business success is off of something different, like design or something. But then they've started that as a side hustle and I'm like, "I don't like it." 

Delores Naskrent: No. "

Kaylie Edwards: I don't like that's what you're doing." But yeah, it ... I have to put my opinions as well in kind of a category where there was point where it was very exciting seeing AI, and I like the idea of the coloring books, like creating the coloring books from, if you had a piece of art and then it was able to turn it into a design.

But then people are generating full coloring books with AI, and a lot of it you could tell. And, Amazon's flooded with it. And - yes, it's easy. Yes, it's quick. But then you gotta look at what about the people that have spent hours, days, [00:29:00] weeks, months creating a coloring book by hand?

And then you think- It where's the trade-off and it, I, yeah, it does upset me a bit. 

But there is a an interesting visual response happening at the same time, brand strategist and creative director Zach Hefner recently posted a reel called Why Naïve Design Is Becoming the Answer to AI.

We should explain that term, naïve design. It doesn't mean the designer is naïve or unskilled. It describes work that looks deliberately simple, childlike, hand-drawn, or imperfect. Wobbly lines, uneven lettering, doodles, layouts that do not feel mechanically polished 

Delores Naskrent: Yeah. Artists will know exactly what we mean by that.

I also call it primitive, and it doesn't mean it's primitive. It just means that it has the character of being loose and spontaneous and still it involves a real tremendous amount of judgment. [00:30:00] Personally, I try to create artwork that is a little bit lively. It might be a little bit wobbly here and there.

It's just because I want people to look at it and know that I did the artwork and it personally, I think, gives it a lot more personality. 

Kaylie Edwards: Yeah. I don't like highly polished stuff, and I think a lot of people are getting very savvy to that now. And Zach points out to brands such as Oatly who is like the oat milk brand, I used to get them all the time, and whose identity uses hand-drawn looking typography, playful copy, and deliberately irregular layouts.

And another brand called Who Gives a Crap, it's the toilet roll company, also uses playful illustration, bold colors, and less corporate visual voice. 

Delores Naskrent: And I find Anthropic especially interesting because it is actually an AI company. Its [00:31:00] design partner, Geist, created this very human illustration style to make some fairly technical ideas feel warmer and easier to understand.

These are done with simple lines, hands, and these wonderfully approachable drawings softening what would otherwise feel like a really cold or complicated subject. 

Kaylie Edwards: Yeah. We should not claim those companies adopted the style only because of generative AI. Oatly has been around for years, and look predates the current boom.

But Zach's wider point is useful when perfectly smooth imagery can be generated in seconds with AI, visible human irregularity becomes a signal of warmth and character. 

Delores Naskrent: And I think that connects so beautifully to handmade work. A little wobble in the line, an unexpected color registration shift, where you got the two colors off each other a little bit.

Maybe pencil markings [00:32:00] or personally I use a lot of just something called a dry ink, which looks a little bit bumpy, has a little bit of pockmarks in it. I change the pressure of my digital brushes so that you can tell when you look at the lines that a real person was present and making those decisions.

I can make a perfect circle in the software. I can do that very easily, but I rather draw the circle myself, and if it's a tiny little bit off, oh

Kaylie Edwards: yeah. And that's what makes it human, isn't it? And what makes it approachable. A 2025 study found that people responded more positively and reported a higher intention to buy art-infused products described as human designed rather than AI designed.

Perceived emotional involvement helped explain the difference. 

Delores Naskrent: Yeah. I don't think that means that we should start adding fake mistakes just to make something look handmade. That [00:33:00] kind of becomes another marketing gimmick all on its own. But maybe we just don't need to polish every lovely little trace of ourselves out of the work either

Kaylie Edwards: Yeah, like I was working on an artwork piece in Affinity Designer and I was- it took me three hours to do this little illustration, and it was of a monkey like holding balloons and things. And I kept going back and then trying to find all the lines where I needed to where the limbs go.

And it was like the little lines and I was like zooming in really far so I could see and then rub it out and then zoom it back out. And I was like, "Why am I doing so much when I could just be spending a lot less time on this?" And just it doesn't matter for some of the mistakes that you forget. 

Delores Naskrent: Yeah, like little marks or small variations, do you notice in your own work?

I personally don't. And in my students' work, I think that's what gives it character rather than making it wrong, personally. 

Kaylie Edwards: Yeah, and [00:34:00] human-made does not mean technology-free. Like digital artists still make hundreds of choices in the software. A Procreate brush can leave a human mark. The question is where the authorship and the judgment live.

And are you more drawn to a piece when you can see sketches, layers, offcuts, tests, and changed direction? Or do you still expect hu- handmade work to look flawless? That is the question to you. And this is also playing out on Etsy. Sellers have reported Etsy testing or automatically applying a physical items filter in searches now.

So that can make digital downloads disappear until the shopper deliberately chooses digital items in the search filters. We could not find an official announcement saying every search now works this way. It may be a test or vary by search and user 

Delores Naskrent: Yeah, if that becomes standard, I can see how it might help some physical handmade sellers in kind of a very broad search [00:35:00] on the platform.

But it could be a real problem for digital artists selling Procreate brushes, like myself, or Affinity assets or SVGs. Let's not forget SVGs and the whole- Yeah ... industry of laser cutting and all of 

Kaylie Edwards: those- Cutting machines, yeah ... 

Delores Naskrent: people buy SVG patterns and they create a product. There's also the creation of repeat patterns for either fabrics or principle- printables or scrapbooking.

Yeah. All of that stuff. What is not acceptable and what is acceptable? 

Kaylie Edwards: Yeah, and digital does not mean less creative either. Like a hand-drawn brush set or a repeat pattern can involve enormous skill and time even though the customer downloads a file. 

Delores Naskrent: Absolutely. And then of course, there's Etsy's rules about computerized tools.

So if you use a Cricut or a laser cutter or any sort of CNC machine or a 3D printer to make the physical [00:36:00] product, Etsy says the design itself needs to be your original work. So you can't buy a pattern, technically, this is what they're saying, is you can't buy the pattern and then cut out the product and sell it as your own product.

So that- Yeah ... I found that very interesting. 

Kaylie Edwards: Yeah, so they're cracking down on that now, so that rule is real, but it was not a new August 2026 crackdown like some of these updates have been. It's Etsy's current creativity standards page, and it's dated since June 2025. So a commercial license can give somebody permission to use a purchased SVG or clip art element.

It does not automatically make it that element the seller's original design, so you can't use it. So anything, even in the Cricut Design Space can't be used unless it has been altered by yourself with other ones. So let's say you built up like a scenery with elements from the Cricut Design Space, you could then [00:37:00] say that's yours because you've put it together.

But yeah, if it's not your original then you wouldn't be able to sell it on Etsy. 

Delores Naskrent: Yeah, I think that's an important difference too that we should talk about. Like the difference between using a properly licensed element as one small part of a new design, and then the difference of taking somebody's ready-made design and putting it on a product almost exactly as is and calling that original.

I think there's a lot of people who just 100% misunderstand it They think if they've used that pattern but they're the ones who've created the item, the cut wood or cut metal or cut glass project, that it's original.

Kaylie Edwards: Yeah. And Etsy's AI policy is more nuanced than AI clip art is banned. Etsy still permits disclosed seller promoted AI creations, but prohibits AI prompt bundles and says digital items must be made or [00:38:00] designed by the seller.

Delores Naskrent: We don't wanna frighten everybody. I don't want you guys all scared because announcing some blanket ban that doesn't actually exist. But generic image packs with very little creative direction, especially when AI use hasn't been disclosed, they're definitely coming under a lot more scrutiny around originality and trust.

Kaylie Edwards: Yeah, and there's been shops that have been closed down or had their listings taken down because of it of late. And then there's new platforms that are responding to this. Fybe, so F-Y-B-E, grew from fiber arts focus and is developing a multi-vendor arts marketplace at the moment that says no generative AI and no drop shipping.

So potential vendors can register interest and should be ready to show how their work is made. So this is like a Etsy alternative, but is obviously they're not selling as of yet. It's in beta [00:39:00] at the moment. 

Delores Naskrent: Yeah. It isn't as established as Etsy at the moment, but a new marketplace still has a huge job of attracting enough buyers.

I do think that its very existence though shows that there are people who are looking for spaces where human made is really the part of the promise. It's not just, what's happened with Etsy, which did start as a handmade marketplace. Unfortunately has gone in so many other different directions that it has lost its footing, in my opinion, and now other marketplaces can spring up, like this one, and replace at least that part of it that used to exist for Etsy people.

Kaylie Edwards: Yeah. Definitely we're gonna be seeing probably more and more of this as Etsy moved away from that direction obviously exchanging hands and what have [00:40:00] you. But question for listeners, would a no generative AI marketplace make you more confident as a buyer or seller? And what proof of process would feel fair without creating another exhausting burden for genuine artists?

And if companies want transparency without fueling more hostility, there are practical changes they could make. First, stop using one broad, vague public label on people's work and tell people whether something was fully generated, partly generated, or only edited with an automated tool 

Delores Naskrent: Yeah. And second, please tell the artist exactly why the label appeared.

Was it because they used the background remover in Canva? Was it because they created a generative fill in Photoshop? Something possibly embedded in the file, or did the creator disclose that information themselves? I don't think we can fix or explain something when we don't even know what caused it. I know people who have had their shops shut down [00:41:00] and absolutely no explanation given.

Kaylie Edwards: Yeah. And third, offer a meaningful appeal before a mistaken label damages a professional reputation. 

Delores Naskrent: Yep. And fourth, give us the option of a truly non-generative workspace. Let an artist switch those features off and export a clear record showing which tools were used and which tools were not used. 

Kaylie Edwards: Yeah.

And fifth, test these systems with working artists, authors, creatives, photographers, handmade sellers before rolling them out widely like they have done. 

Delores Naskrent: Yeah. And why not let the artist add one simple process note beside the label? For example, original watercolor, background removed digitally. Now, I think that would be helpful.

AI info doesn't really tell the customer anything useful at all. Is that watercolor any less original because [00:42:00] the textured paper background was removed? Come on. 

Kaylie Edwards: Yeah, I know. It's just silly. The creative community has a part to play too. I have seen artists and handmade discussions on threads become very polarizing, including some really nasty and blanket anti-AI responses to some genuine artists and creatives, and that is my observation of the community conversation at the moment, not a measured survey, but it shows how raw this feeling has become.

Delores Naskrent: Yeah. People are scared, and for very understandable reasons. Artists are worried that their work has been used without permission. I've seen that a lot. Authors are watching marketplaces fill with low-effort books. I can see that being pretty scary for authors. And designers are wondering whether their jobs will still be there.

Those fears are absolutely real. 

Kaylie Edwards: And valid, and other [00:43:00] solo creators are frightened that one automated edit or one marketing graphic will lead people to decide they are dishonest. I have had a meltdown pretty much this last year because , I was penalized for using, AI for my marketing graphics, because I teach artists and creatives how to market.

Because of that I got penalized for it. And yeah, I did have a bit of a meltdown and almost burnt down my business because of it. And I think that's frightening. Yeah. And it shouldn't be the standard to just scold somebody for using an AI generated marketing piece. 

Delores Naskrent: Yeah.

Being kind doesn't mean that absolutely anything goes.

We can still challenge someone who generates an image and then sells it as hand drawn. We can still ask people to disclose what they used. We can object to artwork being used for training without consent, and to paid creative work being replaced. I think those are all valid things for people [00:44:00] to do, but be kind.

Kaylie Edwards: Yeah. But we can also ask one question before we attack as well. "Can you tell me about your process?" Or, "Why have you used this?" 

Delores Naskrent: Yeah, that one question gives the person a chance to explain. Maybe they made every bit of the artwork and used one automated edit. Maybe they used a generated image for a promotion and are perfectly willing to tell you why, or maybe they really are misleading people, but at least let

then your judgment or our judgment is based on what actually happened, not just what is being said in a blanket sort of a way. 

Kaylie Edwards: Yeah. And for creators, sharing process can be an invitation rather than a defense. So sketches, layers, tests, imperfect versions, materials, and the story behind the final work. 

Delores Naskrent: I don't think artists should have to spend every day proving that they're [00:45:00] innocent, but I love sharing the process.

I think it can be a wonderful thing because of the experiments, the decisions, and even the bits that didn't work. Those are all part of what makes human creativity so interesting. 

Kaylie Edwards: Yeah. And the industry is trying to create rules and labels while the technology, the law, and even the language are still moving and trying to catch up.

Delores Naskrent: So I think we need much more accuracy from the companies creating these labels and a little bit more patience with one another while we all find our way through this.

A label should provide context. It shouldn't feel like an accusation. 

Kaylie Edwards: Yes, and the rise of generated content does not automatically make human creativity less valuable. 

Delores Naskrent: No. In fact, it can make the choices, the marks, and the stories of real makers even more precious. So keep the sketch. Keep that funny little [00:46:00] line that gives your piece its life.

Let people see some of the process and know that a human being was there. 

Kaylie Edwards: A label should begin a question, not finish a judgment. 

Delores Naskrent: Yeah. We would really love to hear what you think about all this. Have you ever had something you created labeled as AI? Or have you seen that label on someone else's work and realized that it changed the way you felt about it?

Kaylie Edwards: And where do you draw the line between assistance and generation, especially when AI is used for marketing rather than the artwork being sold? 

Delores Naskrent: Yeah, come and tell us about your experience and let's please keep the conversation kind. Of the sources we've mentioned today into the show notes as well.

Kaylie Edwards: Yeah. If this episode helped you think about the transition in a more nuanced way, share it with another artist, maker, author, or small business owner and follow the Creative Juggle Joy podcast. 

Delores Naskrent: [00:47:00] Until next time, keep creating, keep juggling, and above all, keep finding joy in the process