IdentityGuard: Context-Aware Restriction and Provenance for Personalized Synthesis
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866915867807186944 |
|---|---|
| author | Zhang, Lingyun Xie, Yu Chen, Ping |
| author_facet | Zhang, Lingyun Xie, Yu Chen, Ping |
| contents | The nature of personalized text-to-image models poses a unique safety challenge that generic context-blind methods are ill-equipped to handle. Such global filters create a dilemma: to prevent misuse, they are forced to damage the model's broader utility by erasing concepts entirely, causing unacceptable collateral damage.Our work presents a more precisely targeted approach, built on the principle that security should be as context-aware as the threat itself, intrinsically bound to the personalized concept. We present IDENTITYGUARD, which realizes this principle through a conditional restriction that blocks harmful content only when combined with the personalized identity, and a concept-specific watermark for precise traceability. Experiments show our approach prevents misuse while preserving the model's utility and enabling robust traceability. By moving beyond blunt, global filters, our work demonstrates a more effective and responsible path toward AI safety. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_15679 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | IdentityGuard: Context-Aware Restriction and Provenance for Personalized Synthesis Zhang, Lingyun Xie, Yu Chen, Ping Cryptography and Security Artificial Intelligence Computer Vision and Pattern Recognition The nature of personalized text-to-image models poses a unique safety challenge that generic context-blind methods are ill-equipped to handle. Such global filters create a dilemma: to prevent misuse, they are forced to damage the model's broader utility by erasing concepts entirely, causing unacceptable collateral damage.Our work presents a more precisely targeted approach, built on the principle that security should be as context-aware as the threat itself, intrinsically bound to the personalized concept. We present IDENTITYGUARD, which realizes this principle through a conditional restriction that blocks harmful content only when combined with the personalized identity, and a concept-specific watermark for precise traceability. Experiments show our approach prevents misuse while preserving the model's utility and enabling robust traceability. By moving beyond blunt, global filters, our work demonstrates a more effective and responsible path toward AI safety. |
| title | IdentityGuard: Context-Aware Restriction and Provenance for Personalized Synthesis |
| topic | Cryptography and Security Artificial Intelligence Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.15679 |