EUVIMEDEUVIMED entdecken
Crossref

iSeal: Encrypted Fingerprinting for Reliable LLM Ownership Verification

Zixun Xiong, Gaoyi Wu, Qingyang Yu, Mingyu Derek Ma, Lingfeng Yao, Miao Pan, Xiaojiang Du, Hao Wang

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (42) · 35940-35948

DOI: 10.1609/aaai.v40i42.40909 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Zixun Xiong und 7 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Given the high cost of large language model (LLM) training from scratch, safeguarding LLM intellectual property (IP) becomes increasingly crucial. As the standard paradigm for IP ownership verification, LLM fingerprinting thus plays a vital role in addressing this challenge. Existing LLM fingerprinting methods verify ownership by extracting or injecting model-specific features. However, they overlook potential attacks during the verification process, leaving them ineffective when the model thief fully controls the LLM's inference process. In such settings, attackers may share prompt-response pairs to enable fingerprint unlearning, or manipulate outputs to evade exact-match verification. We propose iSeal, the first fingerprinting method designed for reliable verification when the model thief controls the suspected LLM in an end-to-end manner. It injects unique features into both the model and an external module, reinforced by an error-correction mechanism and a similarity-based verification strategy. These components are resistant to verification-time attacks, including collusion-based fingerprint unlearning and response manipulation, backed by both theoretical analysis and empirical results. iSeal achieves 100% Fingerprint Success Rate (FSR) on 12 LLMs against more than 10 attacks, while baselines fail under unlearning and response manipulations.

Zitieren

APA

Xiong, Z., Wu, G., Yu, Q., Ma, M. D., Yao, L., Pan, M., Du, X., & Wang, H. (2026). iSeal: Encrypted Fingerprinting for Reliable LLM Ownership Verification. Proceedings of the AAAI Conference on Artificial Intelligence, 40(42), 35940-35948. https://doi.org/10.1609/aaai.v40i42.40909

Vancouver

Xiong Z, Wu G, Yu Q, Ma MD, Yao L, Pan M, et al. iSeal: Encrypted Fingerprinting for Reliable LLM Ownership Verification. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(42):35940-35948. doi:10.1609/aaai.v40i42.40909

BibTeX
@article{Xiong2026,
  author = {Xiong, Zixun and Wu, Gaoyi and Yu, Qingyang and Ma, Mingyu Derek and Yao, Lingfeng and Pan, Miao and Du, Xiaojiang and Wang, Hao},
  title = {iSeal: Encrypted Fingerprinting for Reliable LLM Ownership Verification},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {42},
  pages = {35940--35948},
  doi = {10.1609/aaai.v40i42.40909}
}
Kostenlos starten →