Artists Take up Nightshade Data Poisoning Tool, Fight Back Against Generative AI
Nightshade has gained over 250,000 downloads in its first week after release.
- Computer scientists at the University of Chicago have released a tool called Nightshade that ‘poisons’ digital artwork against scraping by AI models.
- The tool is gaining traction as a copyright protection tool and will likely impact generative AI products such as Midjourney, DALL-E, and Stable Diffusion.
Nightshade, a free tool created by the University of Chicago, has been gaining popularity among artists for protecting digital artwork from AI scraping operations. The tool essentially makes image data useless for AI models that use such data for training. Nightshade makes invisible changes to pixels in an image that makes reproduction by AI unpredictable.
The development is a key move in the growing conflict by content creators against artificial intelligence companies that use web scraping to use art data for AI models without permission. Adopting the tool will likely adversely affect image-generation AI models from Stable Diffusion, DALL-E, and Midjourney.
See More: Google Fixes 17 Bugs in Chrome, Brings GenAI to the Popular Web Browser
AI companies such as Google and OpenAI are already facing numerous lawsuit challenges from artists and publishers for using copyrighted material without permission or any form of compensation. Ben Zhao, the leader of the team that created Nightshade, hopes to tip the balance back toward the artists. Nightshade has already gained 250,000 downloads in the week since its release.
The University of Chicago team also created another tool that allows artists to hide their signature styles from being scraped by AI companies. The tool called Glaze changes how machine learning models view an image, making the end result different from the original image. Since the tools are set to be free and open source, they will likely gain further popularity soon.
What do you think about the potential threats from AI tech? Let us know your thoughts on LinkedIn, X, or Facebook. We’d love to hear from you!
Image source: Shutterstock