Real-Time Detection of Fake AI-Generated Videos Using Deep Learning and Big Data Analytics: A Scalable Cyber Security Strategy

Abstract

The popular production of synthetic media that are of high quality and highly realistic, also referred to as deepfakes, is due to the rapid development of artificial intelligence (AI) and generative models. Although these technologies bring novel opportunities in the entertainment, education, and digital media production domain, they also present a big threat to cybersecurity, privacy, and social trust. The misinformation campaign, identity scams, political influence, and a fraud of money can be conducted with the help of the fake AI-generated videos. Therefore, timely deepfake video detection has emerged as a serious field of research in digital forensics and cybersecurity. The following review paper will present a plethora of current methods of identifying AI-produced fake videos with the help of deep learning and big data analysis. It investigates many types of deep learning frameworks like convolutional neural networks (CNNs), recurrent neural networks (RNNs), generative adversarial network (GAN) detection models, and transformer-based ones. The research also examines how the big data models like Hadoop, Spark and distributed cloud computing have facilitated scalability and real-time detection systems. Additionally, the paper examines benchmark data, measures of evaluation, detection issues, and future directions of research in deepfake detection. The article has identified the use of artificial intelligence coupled with big data infrastructure as essential in developing scalable cybersecurity measures that can process large quantities of multimedia data that are generated in the digital platform. The results propose that spatial, temporal, and behavioral hybrid models are more accurate in the detection. Also, the newly developed technologies explainable AI, federated learning, and blockchain are likely to make the deepfake detection systems even more reliable. The scope of this review will be summarized as offering researchers and cybersecurity practitioners a summarized view of the deepfake detectors and the advancement of real-time scalable security frameworks.

Introduction

Recent advances in artificial intelligence, particularly generative models like GANs and autoencoders, have enabled the creation of highly realistic synthetic media, including deepfake videos. These manipulated videos can convincingly imitate real people, posing risks in cybersecurity, journalism, politics, and social media. Traditional forensic methods struggle to detect such AI-generated media due to sophisticated facial expressions, lighting, and motion patterns.

Deep learning-based detection has become essential. CNNs analyze spatial features of frames, RNNs capture temporal sequences like facial movements, and transformers detect long-range dependencies for enhanced accuracy. Lightweight approaches using motion vector artifacts in compressed video formats (e.g., H.264) enable scalable, real-time detection suitable for social media and streaming platforms.

Big data analytics frameworks like Hadoop, Apache Spark, and cloud computing allow processing large volumes of video data efficiently, integrating deep learning models for real-time detection. Public datasets such as FaceForensics++, DFDC, Celeb-DF, and DeeperForensics are widely used to train and evaluate models. Effective deepfake detection systems combine data collection, preprocessing, deep learning analysis, big data engines, and real-time alert modules to identify manipulated videos at scale.

Conclusion

The Deepfake technology is one of the most relevant cybersecurity threats in the digital age. Artificial intelligence generated videos are becoming real, and this has become dangerous to political stability, financial security, and trust among the population. These videos are hard to detect with a single model of deep learning unless those models are scaled using large infrastructures of big data. The current review paper has analyzed the different deep learning methods that are applied to deepfake detection and these methods are CNNs, RNNs, and transformer-based models. The use of big data analytics to facilitate scalable detection systems had also been talked about. The paper has also reviewed benchmark datasets, detection architectures and some upcoming technologies that may be employed to improve cybersecurity strategies. The results indicate that hybrid detection models involving spatial and temporal analysis are more accurate as compared to single models. In addition, the detection systems can be more reliable and transparent by incorporating explainable AI, federated learning, and blockchain technologies. Since the impact of deepfake technology is still evolving, researchers and cybersecurity experts should devise scalable and responsive solutions in countering the use of AI-generated fake news. The future ways of detecting should be centered on real-time processing, multimodal analysis, and collaborative security frameworks to adequately deal with the increasing threat of fake AI-generated videos.

References

[1] F. Abbas and A. Taeihagh, “Unmasking deepfakes: A systematic review of deepfake detection and generation techniques using artificial intelligence,” 2024. doi: 10.1016/j.eswa.2024.124260. [2] M. Alrashoud, “Deepfake video detection methods, approaches, and challenges,” 2025. doi: 10.1016/j.aej.2025.04.007. [3] L. Stroebel, M. Llewellyn, T. Hartley, T. S. Ip, and M. Ahmed, “A systematic literature review on the effectiveness of deepfake detection techniques,” 2023. doi: 10.1080/23742917.2023.2192888. [4] G. Petmezas, V. Vanian, K. Konstantoudakis, E. E. I. Almaloglou, and D. Zarpalas, “Video deepfake detection using a hybrid CNN-LSTM-Transformer model for identity verification,” Multimed. Tools Appl., 2025, doi: 10.1007/s11042-024-20548-6. [5] N. K. Sagar and S. Arukonda, “A Novel CNN-LSTM Approach for Robust Deepfake Detection,” in Procedia Computer Science, 2025. doi: 10.1016/j.procs.2025.04.436. [6] A. H. Soudy et al., “Deepfake detection using convolutional vision transformers and convolutional neural networks,” Neural Comput. Appl., 2024, doi: 10.1007/s00521-024-10181-7. [7] S. Yadav and S. S. Mangalampalli, “Deepfake defense: Combining spatial and temporal cues with CNN–BiLSTM–transformer architecture,” PLoS One, 2025, doi: 10.1371/journal.pone.0334980. [8] Vishal Manishbhai Patel and Dr. Sheshang Degadwala, “Deepfake Detection Using Convolutional Neural Networks and LSTM Modelling,” Int. J. Sci. Res. Sci. Technol., 2025, doi: 10.32628/ijsrst2512361. [9] A. G. Singh and P. Sharma, “A Hybrid Deep Learning Framework for Robust Deepfake Detection Using CNN, LSTM, and Vision Transformers,” 2024. [10] W. H. Abir et al., “Detecting Deepfake Images Using Deep Learning Techniques and Explainable AI Methods,” Intell. Autom. Soft Comput., 2023, doi: 10.32604/iasc.2023.029653. [11] Y. Zhang, Q. Li, Z. Yu, and L. Shen, “Distilled transformers with locally enhanced global representations for face forgery detection,” Pattern Recognit., 2025, doi: 10.1016/j.patcog.2024.111253. [12] E. M. Sathwik Reddy, A. Pavan Kumar, and P. Swetha, “Deepfake video detection using CNN and RNN with OPTICAL FLOW features,” in 2024 IEEE International Students’ Conference on Electrical, Electronics and Computer Science, SCEECS 2024, 2024. doi: 10.1109/SCEECS61402.2024.10482344. [13] D. Shao, N. Wang, P. Chen, Y. Liu, and L. Lin, “A Novel Video Compression Approach Based on Two-Stage Learning,” Entropy, 2024, doi: 10.3390/e26121110. [14] D. A. Coccomini, R. Caldelli, F. Falchi, and C. Gennaro, “On the Generalization of Deep Learning Models in Video Deepfake Detection,” J. Imaging, 2023, doi: 10.3390/jimaging9050089. [15] A. Naitali, M. Ridouani, F. Salahdine, and N. Kaabouch, “Deepfake Attacks: Generation, Detection, Datasets, Challenges, and Research Directions,” 2023. doi: 10.3390/computers12100216. [16] E. Temir, “Deepfake: New Era in The Age of Disinformation & End of Reliable Journalism TT - Deepfake: Dezenformasyon Ça??nda Yeni Dönem ve Güvenilir Habercili?in Sonu,” Selçuk ?leti?im, 2020. [17] N. Sandotra and B. Arora, “A comprehensive evaluation of feature-based AI techniques for deepfake detection,” 2024. doi: 10.1007/s00521-023-09288-0. [18] F. Ding, R. Kuang, Y. Zhou, L. Sun, X. Zhu, and G. Zhu, “A survey of Deepfake and related digital forensics,” J. Image Graph., 2024, doi: 10.11834/jig.230088. [19] A. Rossler, D. Cozzolino, L. Verdoliva, C. Riess, J. Thies, and M. Niessner, “FaceForensics++: Learning to detect manipulated facial images,” in Proceedings of the IEEE International Conference on Computer Vision, 2019. doi: 10.1109/ICCV.2019.00009. [20] Y. Li, M. C. Chang, and S. Lyu, “In Ictu Oculi: Exposing AI created fake videos by detecting eye blinking,” in 10th IEEE International Workshop on Information Forensics and Security, WIFS 2018, 2018. doi: 10.1109/WIFS.2018.8630787. [21] H. H. Nguyen, F. Fang, J. Yamagishi, and I. Echizen, “Multi-task Learning for Detecting and Segmenting Manipulated Facial Images and Videos,” in 2019 IEEE 10th International Conference on Biometrics Theory, Applications and Systems, BTAS 2019, 2019. doi: 10.1109/BTAS46853.2019.9185974. [22] B. Batagelj, A. Kronovšek, V. Štruc, and P. Peer, “Robust cross-dataset deepfake detection with multitask self-supervised learning,” ICT Express, 2025, doi: 10.1016/j.icte.2025.02.011. [23] L. Guarnera, O. Giudice, and S. Battiato, “DeepFake detection by analyzing convolutional traces,” in IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, 2020. doi: 10.1109/CVPRW50498.2020.00341. [24] D. S. AC, M. K. KM, P. M, P. Chincholi, P. H B, and R. a, “Experimental Detection of Deep Fake Images Using Face Swap Algorithm.,” SSRN Electron. J., 2025, doi: 10.2139/ssrn.5253609. [25] J. Hu, X. Liao, W. Wang, and Z. Qin, “Detecting Compressed Deepfake Videos in Social Networks Using Frame-Temporality Two-Stream Convolutional Network,” IEEE Trans. Circuits Syst. Video Technol., 2022, doi: 10.1109/TCSVT.2021.3074259.

Copyright

Copyright © 2026 Rajendra Singh, Dr. Preeti Gupta. 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.