Deepfake Attacks on Biometric Systems: Threats, Detection, and Defense Mechanisms – A Systematic Survey

DOI: https://doi.org/10.62517/jike.202604108

Author(s)

Weibo Ye

Affiliation(s)

Faculty of Information Technology, Monash University, Melbourne, Victoria, VIC 3168, Australia

Abstract

The rapid advancement of deepfake technology poses unprecedented security threats to biometric systems. This paper presents a systematic survey on the latest progress of deepfake attacks in biometric authentication. First, we construct a Deepfake Kill Chain framework that systematically describes the complete attack chain from content generation to authentication decision. Second, we classify and compare defense methods across four dimensions: content-level, behavior-level, environment-level, and generative-end interventions, analyzing their applicability, failure modes, and trade-offs between security and usability in different scenarios. Third, we deeply discuss how Shortcut Learning leads to reduced generalization capability of detectors on unknown variants. Finally, we propose a hierarchical and collaborative defense framework and provide concrete deployment recommendations. This survey aims to provide a unified cognitive framework for both academia and industry.

Keywords

Deepfake; Biometric Recognition; Detection Methods; Defense Strategies; Model Trustworthiness

References

[1] Li Junjie, Wang Jianzong, Zhang Xulong, Qian Xiaoyang. Generalization problem in video deepfake detection: methods, challenges and technical progress. Big Data Research, 2025. [2] Chingovska I, dos Anjos A R, Marcel S. Biometrics Evaluation Under Spoofing Attacks. IEEE Transactions on Information Forensics and Security, 2014, 9(12): 2264-2277. [3] Gragnaniello D, Poggi G, Sansone C, et al. An Investigation of Local Descriptors for Biometric Spoofing Detection. IEEE Transactions on Information Forensics and Security, 2015, 10(4): 849-863. [4] Li Junjie, Wang Jianzong, Zhang Xulong, Qian Xiaoyang. Generalization problem in video deepfake detection: methods, challenges and technical progress. Big Data Research, 2025. [5] Wang H, Su L, Zeng H, et al. Anti-spoofing study on palm biometric features. Expert Systems with Applications, 2023, 218: 119546. [6] Yang C Z, Ma J, Wang S, et al. Preventing DeepFake Attacks on Speaker Authentication by Dynamic Lip Movement Analysis. IEEE Transactions on Information Forensics and Security, 2021, 16. [7] Nagothu D, Xu R, Chen Y, et al. DeFakePro: Decentralized Deepfake Attacks Detection Using ENF Authentication. arXiv preprint arXiv:2207.13070, 2022. [8] Dong J, Wang Y, Lai J, et al. Restricted Black-Box Adversarial Attack Against DeepFake Face Swapping. IEEE Transactions on Information Forensics and Security, 2023, 18: 2596-2608. [9] Sahidullah S, et al. Shortcut Learning in Binary Classifier Black Boxes: Applications to Voice Anti-Spoofing and Biometrics. arXiv preprint arXiv:2306.00044, 2023. [10] Du Y, Wang J, Zhang X, et al. Cross-modal Fusion with Distillation for DeepFake Detection. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023. [11] Yao W D, Li P C, Zhao Y, et al. Review of research on face deepfake detection methods. Journal of Image and Graphics, 2025, 30(7): 2343-2363. [12] Li C X. Will deepfake bring a new "epistemic crisis"? Studies in Dialectics of Nature, 2024, 40(8). [13] Wang L. Research on Attack and Defense Techniques for Facial Deepfake. Ph.D. dissertation, Shandong University, 2024. [14] Yao W D, Li P C, Zhao Y, et al. Review of research on face deepfake detection methods. Journal of Image and Graphics, 2025, 30(7): 2343-2363. [15] Li J, Wang J, Zhang X, et al. Generalization problem in video deepfake detection: methods, challenges and technical progress. Big Data Research, 2025. [16] Zhang Z, Zhang X, Qian X, et al. Survey on deepfake audio generation and authentication technology. Big Data, 2025. [17] Wang L. Research on Attack and Defense Techniques for Facial Deepfake. Ph.D. dissertation, Shandong University, 2024. [18] Wu H. Research on Anti-Forensic Adversarial Attack Methods against Deepfake Video Detection. M.S. thesis, UESTC, 2024. [19] Mo Y, Zhang Z, Chen Y. Design and Implementation of Face Anti-spoofing Verification System Based on Big Data Analysis. Modern Computer, 2024, 30(16).