Safer Prompts: Reducing IP Risk in Visual Generative AI
This addresses IP infringement concerns for users of visual generative AI models, but it is incremental as it builds on existing prompt engineering methods.
The paper tackled the problem of intellectual property infringement risks in visual generative AI by evaluating prompt engineering techniques, finding that Chain of Thought Prompting and Task Instruction Prompting significantly reduce similarity between generated images and training data.
Visual Generative AI models have demonstrated remarkable capability in generating high-quality images from simple inputs like text prompts. However, because these models are trained on images from diverse sources, they risk memorizing and reproducing specific content, raising concerns about intellectual property (IP) infringement. Recent advances in prompt engineering offer a cost-effective way to enhance generative AI performance. In this paper, we evaluate the effectiveness of prompt engineering techniques in mitigating IP infringement risks in image generation. Our findings show that Chain of Thought Prompting and Task Instruction Prompting significantly reduce the similarity between generated images and the training data of diffusion models, thereby lowering the risk of IP infringement.