Sweep your site

Have you been scraped? Make it harder for A.I. to train on your art in two steps - scan & clean.


Enter the URL of a platform you're using. See if your images are being referenced in the LAION dataset.


Guides on how to de-list your images from the LAION dataset

Why should you use ArtShield Sweeper?

Find your artwork in public AI training datasets, so you can reclaim your art and implement preventative measures.

How the Sweeper Works

Because public dataset providers don't want to be liable for distribting copyrighted material, they distribute links to images instead of images themselves. URLs also require less storage space than images.

The Sweeper tool will reference a user’s store page domain(url) that’s been inputted into the search bar. Doing this manually as an individual would take way too long as there are millions of urls stored on these databases, but Sweeper is helping you with the process.

There are many public datasets that have scraped artists' work

The ArtShield Sweeper tool cross-referencings the your submitted platform with these public datasets. Currently we check the LAION 5B and 400M datasets. We will also support more public datasets as they are made known.

Once you know which image links are in datasets, you have the power to break the URL

ArtShield Sweeper can help you find links for your art in training datasets, so you can break the link to no longer point to your image. (After you sweep, we'll also show you a guide for breaking the links!)

The more broken URLs, the less the future A.I are able to train

We hope that by arming you with this knowledge about which images of yours are in the datasets, you are able to fight legal battles, remove your art from the collection, and poison the readily available public datasets.


Currently Supported Platforms

Platforms In Development


Will you support ___ platform in the future?

What is an AI training dataset?

What are LAION2B-en, LAION2B-multi, and LAION1B-nolang? How are they different from LAION-5B?

Why can't I just search the dataset myself?

How is ArtShield Sweeper different from HaveIBeenTrained?

Should we also use HaveIBeenTrained?

Why are you building an opt-out solution instead of an opt-in solution?

How can reuploading an image and breaking existing links impact my SEO?