







Images I found on Twitter, “free adoptable raffle” characters. Is this real? (I doubt it is)
The Twitter account got nuked or deactivated by the owner, so I won’t link the source.
When I first saw these, I didn‘t look at them too closely and entered the raffle, before reverting it after a VTuber artist whose name I can’t remember pointed out that these were likely AI generated in the replies.
Also, I would like to note that I am an artist myself, I know (somewhat) how to spot AI generated images, I know what I’m talking about.
I made a thread dissecting these images on my private twitter account, so let me repeat the things I said there and add a few things:
— New artist shows up out of nowhere with no accounts connecting to the username (which I won’t list as some people would likely be compelled to harass them, which would only make things worse!) and no/very small backlog of past art
There was an “old art” post that got no attention, but that also could have been generated and/or traced from AI. I don’t count a single image claiming to be old art as proof. There was also a $70 USD Chibi adoptable that got some attention, but not much.
The images 4, 5 and 6 are me dissecting the photos, so let me explain why i Red Outlined it:
GAL 1
The first thing that stuck out to me was the hair. I am an artist myself who has seen many AI images to train myself on how to spot them, and let me tell you: AI does NOT know how hair works. The hair is a melty mess, with looping hairstrands galore (though you can argue that her hair is simply curly).
The bow on the left side of the image looks like its missing the other half, or its going behind her hair, which wouldn’t make much sense.
One hand has 3 fingers when the other has 4. This error is present in Gal 2 as well.
Also, the clover is. Going over a line??? I don’t think it should be doing that. Also, the spacing of the clovers is inconsistent.
GAL 2
3 fingers on one hand, 4 on the other.
Ruffles on shorts are melting/make no sense.
I didn’t redline it because of it possibly being intentional, but the stockings stripes aren’t symmetrical at all.
Eyelashes melting into hair somewhat on the left side (not outlined because I didn’t immediately notice it until I made the face closeups)
The pink flower things don’t make any sense to be here, and the tiny one is MELTING.
Inconsistent detailing on the shoes (one has a buckle while the other doesn’t, bows aren’t equally detailed)
GAL 3
Legs are broken! So very broken! She’s in pain, save her!
One eyebrow has 2 lines when the other has 1.
Her tail disappears behind her leg, and her sweater collar disappears behind her necklace. Generative AI has an issue with object permanence.
Also, she has heavier shadows than the other 2 girls.
The second to last image is to compare the faces. Tell me: for reasons other than the brush, do these look like one person drew these? With the multiple eyebrows and how different the eye rendering is, I’d have to say no. Not one person would draw these eyes all so differently.
The final image is a few of the faces I have drawn recently, about 3 weeks ago? It exists to compare the two, to compare AI to human. You can see the faces have CONSISTENCY. Outside of the eye shines. You can tell “hey someone made all of those!”, which you can’t with the likely AI ones.
Also, I suspect a LoRA was used in making this, given the slightly consistent art style. If someone could tell me which artist the hypothetical LoRA was trained off of, please inform me! I would greatly appreciate it, because this is a very charming art style!
Anyways, does anyone agree that these are AI? Or am I just nitpicking??