We all know that image models like OpenAI’s Dall-E can be jailbroken to generate violent, disturbing, and offensive images.
