Adversarial Perturbations Cannot Reliably Protect Artists From Generative AI

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hönig, Robert, Rando, Javier, Carlini, Nicholas, Tramèr, Florian
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929708450447360
author Hönig, Robert
Rando, Javier
Carlini, Nicholas
Tramèr, Florian
author_facet Hönig, Robert
Rando, Javier
Carlini, Nicholas
Tramèr, Florian
contents Artists are increasingly concerned about advancements in image generation models that can closely replicate their unique artistic styles. In response, several protection tools against style mimicry have been developed that incorporate small adversarial perturbations into artworks published online. In this work, we evaluate the effectiveness of popular protections -- with millions of downloads -- and show they only provide a false sense of security. We find that low-effort and "off-the-shelf" techniques, such as image upscaling, are sufficient to create robust mimicry methods that significantly degrade existing protections. Through a user study, we demonstrate that all existing protections can be easily bypassed, leaving artists vulnerable to style mimicry. We caution that tools based on adversarial perturbations cannot reliably protect artists from the misuse of generative AI, and urge the development of alternative non-technological solutions.
format Preprint
id arxiv_https___arxiv_org_abs_2406_12027
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Adversarial Perturbations Cannot Reliably Protect Artists From Generative AI
Hönig, Robert
Rando, Javier
Carlini, Nicholas
Tramèr, Florian
Cryptography and Security
Artists are increasingly concerned about advancements in image generation models that can closely replicate their unique artistic styles. In response, several protection tools against style mimicry have been developed that incorporate small adversarial perturbations into artworks published online. In this work, we evaluate the effectiveness of popular protections -- with millions of downloads -- and show they only provide a false sense of security. We find that low-effort and "off-the-shelf" techniques, such as image upscaling, are sufficient to create robust mimicry methods that significantly degrade existing protections. Through a user study, we demonstrate that all existing protections can be easily bypassed, leaving artists vulnerable to style mimicry. We caution that tools based on adversarial perturbations cannot reliably protect artists from the misuse of generative AI, and urge the development of alternative non-technological solutions.
title Adversarial Perturbations Cannot Reliably Protect Artists From Generative AI
topic Cryptography and Security
url https://arxiv.org/abs/2406.12027