GoodDiffusion: Proactive Copyright Protection for Diffusion Bridge Models via Learnable Sample-specific Signatures

This paper tackles the challenging problem of developing a proactive copyright protection mechanism that cuts off unauthorized use of diffusion bridge models. Existing studies largely fall into post-hoc attribution (e.g., watermarking and fingerprinting) or degradation-only defenses, which offer only indirect and limited preventive effect. We therefore propose GoodDiffusion, inspired by backdoor mechanisms, to enforce model-level use-time control by internalizing authorization into the generative process through a selectively permissive, otherwise closed behavior. Specifically, GoodDiffusion preserves high-quality generation for authorized queries carrying valid signatures, yet refuses to generate for unauthorized inputs. We further empirically show that naive static-signature designs (like conventional backdoor injection) are fundamentally fragile, since a surrogate signature can be efficiently recovered via gradient-based optimization. To strengthen security, we introduce a Learnable Signature Network (LSN) that assigns sample-specific signatures conditioned on each input. This breaks the universality of signatures and prevents a surrogate from transferring across inputs. Extensive experiments validate that GoodDiffusion effectively blocks unauthorized use while maintaining strong generation quality for authorized users.

Modern diffusion models are expensive to train and valuable to their owners, but once a copy of a model is leaked, someone may use it without permission. Thus, the copyright protection of diffusion models is crucial for the development of generative AI. Existing protections often act like watermarks: they can prove misuse after images are made, but they do not stop misuse from happening. We study how to make a diffusion bridge model behave more like a licensed service: no ticket, no ride. Specifically, we propose GoodDiffusion, which gives each authorized input a small learned signature, similar to a per-image ticket, and trains the model to produce high-quality results only when the valid signature is present. If the input is unauthorized, the model returns a warning image to indicate the unauthorized usage. We evaluate this idea on various Image-to-Image tasks, and the results show that GoodDiffusion can effectively block unauthorized use while preserving generation quality for authorized users. This work promotes the copyright protection from the post-hoc tracing to proactive prevention at the generation stage.