GoodDiffusion: Proactive Copyright Protection for Diffusion Bridge Models via Learnable Sample-specific Signatures
2026-06-29 • Cryptography and Security
Cryptography and Security
AI summaryⓘ
The authors address the problem of stopping unauthorized use of diffusion bridge models by creating a new system called GoodDiffusion. Unlike existing methods that try to fix misuse after it happens, GoodDiffusion controls access in real-time by allowing only verified inputs to generate quality outputs and blocking others. They found that simple static keys for authorization can be tricked, so they designed a Learnable Signature Network that creates unique keys based on each input, making it much harder to bypass. Their experiments show this approach blocks unauthorized use without hurting performance for allowed users.
diffusion bridge modelscopyright protectionbackdoor mechanismwatermarkingfingerprintinggradient-based optimizationLearnable Signature Networkauthorizationmodel securitygenerative process
Authors
Shixi Qin, Zhiyong Yang, Shilong Bao, Zitai Wang, Qianqian Xu, Qingming Huang
Abstract
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 effects. 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 theoretically 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.