Malicious use of Deepfakes in IoT environments: a systematic mapping of threats, vectors, and mitigation strategies

• Almeida Italo M. Santos UFS
• Marcos Vinícius de S. Santos UFS
• Rubens de Souza M. Junior UFS

DOI: https://doi.org/10.5753/erbase.2025.13700

Abstract

Context: The growing integration of AI and IoT technologies fuels smart environments but also brings new cybersecurity threats like deepfakes. These synthetic media can deceive facial recognition, voice authentication, and surveillance in connected spaces. Objective: This work systematically maps the malicious use of deepfakes in IoT, pinpointing attack vectors, vulnerable sensors, and mitigation techniques. Method: We conducted a systematic mapping, searching IEEE Xplore and Scopus. From 85 initial studies, we selected 16, analyzing them based on three research questions concerning threats, vulnerabilities, and defense mechanisms. Results: Our analysis shows facial and voice authentication systems are most affected. Key attack vectors include face spoofing, voice cloning, and live video manipulation. Effective detection and mitigation techniques in IoT scenarios involve CNNs, blockchain, GNNs, and multimodal detection (Wi-Fi + video). Conclusion: Deepfakes are a concrete and evolving threat to IoT security. Robust, lightweight detection models, alongside biometric authentication frameworks and cryptographic protection, are crucial for real-time defense in resource-constrained IoT environments.

References

Bethu, S. and Erukala, S. B. (2025). Blockchain-enhanced secure guard: a deep q network framework for robust iot surveillance person detection. Cluster Computing, 28(4). Cited by: 0.

Bethu, S., Trupthi, M., Mandala, S. K., Karimunnisa, S., and Banu, A. (2024). Ai-iot enabled surveillance security: Deepfake detection and person re-identification strategies. International Journal of Advanced Computer Science and Applications, 15(7):1013 – 1022. Cited by: 1; All Open Access, Gold Open Access.

Eidmum, M. Z. A., Muiz, B., and Saha, S. (2025). Enhancing fake face detection with meta-learning and ensemble methods. International Journal of Computing and Digital Systems, 18(1). Cited by: 0.

Fang, X., Liu, J., Chen, Y., Han, J., Ren, K., and Chen, G. (2023). Nowhere to hide: Detecting live video forgery via vision-wifi silhouette correspondence. volume 2023-May. Cited by: 3.

Frolov, D. B., Makhaev, D. D., and Shishkarev, V. V. (2022). Deepfakes and information security issues. In 2022 International Conference on Quality Management, Transport and Information Security, Information Technologies (ITQMIS), pages 147–150.

Javed, A., Malik, K. M., Malik, H., and Irtaza, A. (2022). Voice spoofing detector: A unified anti-spoofing framework. Expert Systems with Applications, 198. Cited by: 29; All Open Access, Bronze Open Access.

Karathanasis, A., Violos, J., and Kompatsiaris, I. (2025). A comparative analysis of compression and transfer learning techniques in deepfake detection models. Mathematics, 13(5). Cited by: 0; All Open Access, Gold Open Access.

Liu, J., Fang, X., Chen, Y., Yuan, J., Yu, G., and Han, J. (2025). Real-time video forgery detection via vision-wifi silhouette correspondence. IEEE Transactions on Mobile Computing, 24(3):1585–1601.

Mitra, A., Bigioi, D., Mohanty, S. P., Corcoran, P., and Kougianos, E. (2022). Iface 1.1: A proof-of-concept of a facial authentication based digital id for smart cities. IEEE Access, 10:71791 – 71804. Cited by: 11; All Open Access, Gold Open Access.

Mitra, A., Mohanty, S. P., Corcoran, P., and Kougianos, E. (2021a). Detection of deep-morphed deepfake images to make robust automatic facial recognition systems. page 149 – 154. Cited by: 12.

Mitra, A., Mohanty, S. P., Corcoran, P., and Kougianos, E. (2021b). iface: A deepfake resilient digital identification framework for smart cities. In 2021 IEEE International Symposium on Smart Electronic Systems (iSES), pages 361–366.

Petersen, K., Vakkalanka, S., and Kuzniarz, L. (2015). Guidelines for conducting systematic mapping studies in software engineering: An update. Information and Software Technology, 64:1–18.

Samrouth, K., El Housseini, P., and Deforges, O. (2025). Siamese network-based detection of deepfake impersonation attacks with a person of interest approach. ACM Transactions on Multimedia Computing, Communications and Applications, 21(3). Cited by: 0.

Sisson, E. S. and Puckett, S. C. (2025). Securing digital media in the age of ai and deepfakes: A hardware-based solution. page 1171 – 1176. Cited by: 0.

Xu, J., Lin, W., Fan, W., Chen, J., Li, K., Liu, X., Xu, G., Yi, S., and Gan, J. (2024). A graph neural network model for live face anti-spoofing detection camera systems. IEEE Internet of Things Journal, 11(15):25720–25730.

Zhang, G., Gao, M., Li, Q., Guo, S., and Jeon, G. (2024). Detecting sequential deepfake manipulation via spectral transformer with pyramid attention in consumer iot. IEEE Transactions on Consumer Electronics, pages 1–1.

Zhou, L., Ma, C., Wang, Z., Zhang, Y., Shi, X., and Wu, L. (2024). Robust frame-level detection for deepfake videos with lightweight bayesian inference weighting. IEEE Internet of Things Journal, 11(7):13018–13028.