A Study on Human-Computer Interaction in Deepfake Detection Systems

Abstract

Human-Computer Interaction (HCI) plays a crucial role in the design of intelligent systems that humans can understand, trust, and effectively use. With the rapid advancement of artificial intelligence, deepfakes have emerged as a significant challenge to digital trust. Deepfakes are synthetic media generated using machine learning techniques such as Generative Adversarial Networks (GANs), capable of producing highly realistic manipulated images, videos, and audio. Although existing deepfake detection systems achieve high ac- curacy under controlled conditions, many prioritize algorithmic performance while neglecting usability, transparency, and human trust. This paper explores the integration of HCI principles into deepfake detection systems. We review existing detection methodologies, analyze cognitive load and explainable artificial intelligence (XAI), and examine ethical implications associated with synthetic media. A user-centered framework is proposed to enhance system usability and trust without compromising detection accuracy. The results highlight the importance of ex- plainability, minimal cognitive load, and ethical interface design in improving real-world adoption of deepfake detection tools.

Introduction

The text examines deepfake detection through the lens of Human–Computer Interaction (HCI), emphasizing that technical accuracy alone is insufficient for real-world adoption. While deep learning techniques such as GANs have enabled realistic synthetic media, they have also intensified risks related to misinformation, identity misuse, and loss of trust in digital communication. Although detection methods in computer vision and multimedia forensics have improved, many fail to address usability, transparency, and user trust.

The literature review highlights the evolution of deepfake detection from visual artifact analysis and frequency-domain methods to advanced multimodal approaches combining audio, video, and temporal cues. However, detection models often suffer performance degradation on unseen data, and users generally struggle to identify deepfakes without assistance. Explainable AI (XAI) has been shown to improve trust, though excessive complexity can increase cognitive load.

To address these gaps, the paper proposes an HCI-centered framework that integrates user-centered interface design, explainable AI, cognitive load optimization, multimodal detection, and ethical safeguards. Key features include adaptive interfaces for different user expertise levels, visual explanations of model decisions, progressive disclosure of information, user feedback loops, and privacy- and bias-aware design.

Comparative analysis shows that while traditional and multimodal detection approaches achieve high accuracy, they offer limited explainability. The proposed HCI-centered approach achieves both high accuracy and high interpretability, leading to greater user trust. The study concludes that combining technical detection methods with HCI principles and human-in-the-loop design is essential to overcoming challenges such as usability trade-offs, ethical concerns, and real-world performance decay in deepfake detection systems.

Conclusion

This paper emphasizes that deepfake detection is not solely a technical problem but a socio-technical challenge requiring human-centered solutions. Integrating HCI principles into detection systems improves usability, transparency, and trust, which are critical for real-world adoption. This study demonstrates that effective deepfake detection extends be- yond algorithmic accuracy and must be approached as a human–computer interaction problem. By combining explainable AI, cognitive load-aware inter- face design, and multimodal detection, systems can better support users in navigating synthetic media. Future work should explore adaptive learning strategies and cross-cultural usability studies to address evolving deepfake threats. Aligning algorithmic performance with human needs is essential for building trustworthy and ethically responsible detection systems.

References

[1] D. Afchar, V. Nozick, J. Yamagishi, and I. Echizen, “MesoNet: A compact facial video forgery detection network,” in Proc. IEEE Int. Workshop on Information Forensics and Security (WIFS), 2018. [2] X. Yang, Y. Li, and S. Lyu, “Exposing deepfakes using inconsistent head poses,” in Proc. IEEE Int. Conf. on Acoustics, Speech and Signal Processing (ICASSP), 2019. [3] S. Agarwal, H. Farid, Y. Gu, M. He, K. Nagano, and H. Li, “Protecting world leaders against deepfakes,” in Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR) Workshops, 2019. [4] P. Korshunov and S. Marcel, “Deepfakes: A new threat to face recog- nition? Assessment and detection,” arXiv preprint arXiv:1904.07399, 2019. [5] R. Tolosana, R. Vera-Rodriguez, J. Fierrez, A. Morales, and J. Ortega- Garcia, “Deepfakes and beyond: A survey of face manipulation and detection,” Information Fusion, vol. 64, pp. 131–148, 2020. [6] Y. Mirsky and W. Lee, “The creation and detection of deepfakes: A survey,” ACM Computing Surveys, vol. 54, no. 1, pp. 1–41, 2021. [7] H. Khalid, S. Woo, and S. Lee, “FakeAVCeleb: A novel audio-video multimodal deepfake dataset,” in Proc. Int. Conf. on Multimedia Re- trieval (ICMR), 2021. [8] G. Gupta, K. Raja, M. Gupta, T. Jan, S. T. Whiteside, and M. Prasad, “A comprehensive review of deepfake detection using advanced machine learning and fusion methods,” Electronics, vol. 13, no. 1, p. 95, 2024. [9] T. P. Nagarhalli, A. Save, S. Patil, and U. Aswalekar, “A comprehensive review of deepfake and its detection techniques,” SSRG International Journal of Electrical and Electronics Engineering, 2024. [10] K. Somoray, D. Miller, and M. Holmes, “Human performance in deep- fake detection: A systematic review,” Human Behavior and Emerging Technologies, 2025. [11] J. Richings, M. Leblanc, I. Groves, and V. Nockles, “Performance decay in deepfake detection: The limitations of training on outdated data,” arXiv preprint arXiv:2511.07009, 2025. [12] A. E. Smith and L. Chen, “Explainable AI and user trust in human- centered security systems,” IEEE Computer, vol. 56, no. 3, pp. 45–53, 2023.

Copyright

Copyright © 2026 Miss. Renuka Khade, Miss. Aishwarya Dharmadhikari, Miss. Priyanka Khandagale. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.