ATTENTIONMultiple sites have been targeted by an AI dataset scraper including artfol.
You can do something about this, detailed instructions are avialable on Paperdemon.com simply look at this post: https://bsky.app/profile/paperdemonarpg.bsky.social/post/3ln4juguvns2g
In case that link doesn’t work: https://www.paperdemon.com/app/g/pdarpg/events/view/994/immediate-action-required-your-art-and-writing-has-been-scraped-and-published-in-an-ai-dataset/1
Filing a takedown request takes just a few minutes. Please repost so everyone can see this and react
Edit: If the link doesnt work for you just mention it in the comments I will dm it to you
Edit2: the artfol team is aware of this, so dont unnecessarily spam them about it
Edit3: official statement and instructions to take down your work from the artfol dev team!!!! https://artfol.app/dev
Edit 4: I got an email from the person who scraped all the data, Artfols dataset DOES NOT include actual art, just metadata. So we are soorta in the clear. Still other datasets from the other affected sites do contain writing and artworks.
This confusion might be my fault, as I misread the dataset description (while I was panicking about PaperDemon, which´s dataset does contain artworks).
Frequently asked questions:
Would deleting my account/art help?
-deleing your artwork or account from the mentioned social media sites wont help, the artworks were already stolen days/weeks ago.
The process is to tedious/ I dont want to disclose my information
-at least report the dataset on huggingface (the site where the dataset was uploaded) for copyright infringement if you can, this will still help in taking down the dataset
Has an AI model been trained already?
-no this dataset only includes the data from (for example artfol), posts, art, tags, etc. This dataset can then be downloaded to train AI models. We can assume that a few people did download it already though.
How can I check if my art is affected?
-if you posted on any of the listed sites you can assume that your posts are included in the dataset/s, since the scraper probably took all the data (posts) on the site. For example the Artfol datast includes over a million posts.
The only way to know for sure is to download the whole dataset and look for your stuff yourself (not recommended, probably takes up a lot of storage, I also dont know how it works)
I want to give up drawing/ posting
-please dont, we dont want to loose any more wonderful artists and creative minds. AI has won if we give up hope. Fighting against this will also give the message that we dont tolerate AI and could potentially aid in laws being made to takle the stealing issue.
To keep your art safe in the future look into glaze/ nightshade (if you have a beefy computer) or cover your art with watermarks or transparent random pictures to confuse an AI that tries to train on your work.
Apr 18, 2025
You can do something about this, detailed instructions are avialable on Paperdemon.com simply look at this post: https://bsky.app/profile/paperdemonarpg.bsky.social/post/3ln4juguvns2g
In case that link doesn’t work: https://www.paperdemon.com/app/g/pdarpg/events/view/994/immediate-action-required-your-art-and-writing-has-been-scraped-and-published-in-an-ai-dataset/1
Filing a takedown request takes just a few minutes. Please repost so everyone can see this and react
Edit: If the link doesnt work for you just mention it in the comments I will dm it to you
Edit2: the artfol team is aware of this, so dont unnecessarily spam them about it
Edit3: official statement and instructions to take down your work from the artfol dev team!!!! https://artfol.app/dev
Edit 4: I got an email from the person who scraped all the data, Artfols dataset DOES NOT include actual art, just metadata. So we are soorta in the clear. Still other datasets from the other affected sites do contain writing and artworks.
This confusion might be my fault, as I misread the dataset description (while I was panicking about PaperDemon, which´s dataset does contain artworks).
Frequently asked questions:
Would deleting my account/art help?
-deleing your artwork or account from the mentioned social media sites wont help, the artworks were already stolen days/weeks ago.
The process is to tedious/ I dont want to disclose my information
-at least report the dataset on huggingface (the site where the dataset was uploaded) for copyright infringement if you can, this will still help in taking down the dataset
Has an AI model been trained already?
-no this dataset only includes the data from (for example artfol), posts, art, tags, etc. This dataset can then be downloaded to train AI models. We can assume that a few people did download it already though.
How can I check if my art is affected?
-if you posted on any of the listed sites you can assume that your posts are included in the dataset/s, since the scraper probably took all the data (posts) on the site. For example the Artfol datast includes over a million posts.
The only way to know for sure is to download the whole dataset and look for your stuff yourself (not recommended, probably takes up a lot of storage, I also dont know how it works)
I want to give up drawing/ posting
-please dont, we dont want to loose any more wonderful artists and creative minds. AI has won if we give up hope. Fighting against this will also give the message that we dont tolerate AI and could potentially aid in laws being made to takle the stealing issue.
To keep your art safe in the future look into glaze/ nightshade (if you have a beefy computer) or cover your art with watermarks or transparent random pictures to confuse an AI that tries to train on your work.
Apr 18, 2025
Comments
FOR THE PEOPLE WHO RECEIVED AN ANSWER
DO NOT let yourself be guilt trip or reconsider the validity of our claims JUST BECAUSE it is metadata and doesn't contain images. METADATA IS INFORMATION. Which also contains LINKS.
I have written a response that I will add in reply to clarify my claim.
Update, we just published instructions, as well as a tool to help you export your art in a csv format, so you can request a takedown of your content! Doing so should also help speed up our request for removal of the whole dataset. Please let me know if you have any issues with this process.
Detailed instructions can be found here: https://artfol.app/dev
@Snakewithblade if you can add this link to your post, that would be great!
Check this out https://huggingface.co/datasets/nyuuzyou/artfol/discussions/91?not-for-all-audiences=true
•
Thank you for this, I'm still using Art Shield and hopefully glaze for my future art before I post it online.
•
The user is totally also against the content policy of hugging face. https://discuss.huggingface.co/faq we may have to report the user to hugging face too. Also here is hugging faces email. legal@huggingface.co
•
Thank you for this post 
once we download the CSV and put it into an email with the template, do we still need to include additional links? Or is it alright to send?
once we download the CSV and put it into an email with the template, do we still need to include additional links? Or is it alright to send?•
Yeah do not fall for it meta data is a pretty huge thing, that is still the info of the pictures
•
FOR THE PEOPLE WHO RECEIVED AN ANSWERDO NOT let yourself be guilt trip or reconsider the validity of our claims JUST BECAUSE it is metadata and doesn't contain images. METADATA IS INFORMATION. Which also contains LINKS.
I have written a response that I will add in reply to clarify my claim.
•
•It's getting harder for young artists,including myself to simply post art in the internet without fearing AI and the world's rising dependance over it.
•
Just finished following the instructions, hopefully they will remove not just mine, but every artist's data from that site... It's just horrible that this is happening right now :(
•
Oh my god... Maybe it's because I disappear from this app for days that I do not know this... Thanks for the links... I will try to read it...
•
Hey sorry if this sounds stupid but where do I send the letter on step 7? Do I send it via direct email to the person along with the CVS?
•
•
Hey! So would every single one of us have to file a report? Or would just the devs being already aware be enough to help us all. Doing the report itself personally just feels like a headache to do…I’m really just tired of AI trying to ruin our genuine hard work.
•
•I got extremely confused at the 6th step, do I have to email the person? If so, how do I email my complaint to this person
•
•How can I manually check if my art is feeding their dataset, as the Google Doc suggests? The CSV is for compiling the information (Title, link) about your own artwork, without information about whether your artwork remains in the dataset or not. I tried to download the dataset directly to my computer, but for some reason unzipping it doesn't work.
•
•Did this have anything to do with the timeout that happened this morning? I remember reading on Mastodon that getting your server scraped can incur massive costs, and it was a contributing factor to the shutdown of one of the most famous servers (botsin.space) I imagine with an image-heavy site like this, that might be a fear too. Is Artfol doing okay financially?
•
•I will be doing all the steps listed as soon as possible, it's disgusting that these frauds are able to steal from people like this, and I have nothing but respect for Artfol for trying to help people to get their content taken down from this model
•
•I’ll definitely do this!! However, I do have a question. once I’m done reporting, will I be able to delete my account on that site? I know it’s for a good and important reason, but I don’t feel comfortable keeping an account that’s related to AI 
•
•
I have filed a copyright infringement report but I am not sure what to do with the CSV thing, it's for removing artworks individually and I don't know which of them got stolen. Is there a way to check? Or is filing a general report good enough? The step-by-step makes it seem like all of the steps need to be completed, and I think I can't do CSV without knowing which ones to send
•
•
Update, we just published instructions, as well as a tool to help you export your art in a csv format, so you can request a takedown of your content! Doing so should also help speed up our request for removal of the whole dataset. Please let me know if you have any issues with this process.Detailed instructions can be found here: https://artfol.app/dev
@Snakewithblade if you can add this link to your post, that would be great!
•
•What kind of monster trains a model without explicit consent? Isn’t this illegal??
And on top of that, the scraper wants all of us to individually call for a takedown. Ridiculous- It should be enough to say “Hey- We didn’t give consent to this-“
I’m fine with AI by itself. Do whatever you want- But you CANNOT (or should not be able to if they are…) legally use someone else’s content without EXPLICIT CONSENT-
And on top of that, the scraper wants all of us to individually call for a takedown. Ridiculous- It should be enough to say “Hey- We didn’t give consent to this-“
I’m fine with AI by itself. Do whatever you want- But you CANNOT (or should not be able to if they are…) legally use someone else’s content without EXPLICIT CONSENT-
•
•I'm trying to send a takedown request however I'm a little confused on the formatting for the file with all the urls. Is it simply just putting the header/title followed by the corresponding url for each of my posts?
•
•
I read through the doc, and I'm confused. how do you file a takedown if your're on Artfol. sorry for the stupid question
•
•