4chan Users Exploit Leaked Stable Diffusion Model for Celebrity Deepfakes
An early leak of the Stable Diffusion image generator allows 4chan users to create non-consensual celebrity deepfakes, highlighting the dangers of open-source AI misuse.
Users on the 4chan message board exploit a leaked version of Stable Diffusion to generate pornographic deepfakes of celebrities. The open-source image generation model, developed by Stability AI and partners, reportedly reaches the platform before its official release date.
This unauthorized distribution allows malicious actors to bypass the model's built-in safety filters. As a result, bad actors quickly use the tool to create explicit, non-consensual images, causing significant harm to the targeted individuals.
This incident highlights the severe risks associated with post-deployment misuse of open-source AI systems. It serves as a prominent example of how malicious actors manipulate accessible AI tools for targeted manipulation and fraud.