Deepfake Technology in Information Warfare: A Multimodal Detection and Automated Counter- Propaganda Framework

References

[1] I. J. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y. Bengio, "Generative adversarial nets," in Proc. 27th Int. Conf. Neural Information Processing Systems (NIPS), vol. 2, 2014, pp. 2672–2680.

[2] A. Rössler, D. Cozzolino, L. Verdoliva, C. Riess, J. Thies, and M. Nießner, "FaceForensics++: Learning to detect manipulated facial images," in Proc. IEEE/CVF Int. Conf. Computer Vision (ICCV), Seoul, South Korea, 2019, pp. 1–11.

[3] R. Tolosana, R. Vera-Rodriguez, J. Fierrez, A. Morales, and J. Ortega-Garcia, "DeepFakes and beyond: A survey of face manipulation and fake detection," Information Fusion, vol. 64, pp. 131–148, 2020.

[4] R. Chesney and D. K. Citron, "Deep fakes: A looming challenge for privacy, democracy, and national security," California Law Review, vol. 107, pp. 1753–1819, 2019.

[5] M. Todisco, X. Wang, V. Vestman, M. Sahidullah, H. Delgado, A. Nautsch, J. Yamagishi, N. Evans, T. Kinnunen, and K. A. Lee, "ASVspoof 2019: Future horizons in spoofed and fake audio detection," in Proc. Interspeech, Graz, Austria, 2019.

[6] S. Vosoughi, D. Roy, and S. Aral, "The spread of true and false news online," Science, vol. 359, no. 6380, pp. 1146–1151, 2018.

[7] S. C. Woolley and P. N. Howard, Eds., Computational Propaganda: Political Parties, Politicians, and Political Manipulation on Social Media. Oxford, U.K.: Oxford Univ. Press, 2018.

[8] C. Hutto and E. Gilbert, "VADER: A parsimonious rule-based model for sentiment analysis of social media text," in Proc. Int. AAAI Conf. Web and Social Media (ICWSM), vol. 8, no. 1, 2014, pp. 216–225.

[9] Y. Liu, M. Ott, N. Goyal, J. Du, M. Joshi, D. Chen, O. Levy, M. Lewis, L. Zettlemoyer, and V. Stoyanov, "RoBERTa: A robustly optimized BERT pretraining approach," arXiv:1907.11692, 2019.

[10] F. Barbieri, J. Camacho-Collados, L. Espinosa-Anke, and L. Neves, "TweetEval: Unified benchmark and comparative evaluation for tweet classification," in Findings of the Assoc. Comput. Linguistics: EMNLP 2020, 2020, pp. 1644–1650.

[11] M. Tan and Q. V. Le, "EfficientNet: Rethinking model scaling for convolutional neural networks," in Proc. 36th Int. Conf. Machine Learning (ICML), Long Beach, CA, USA, 2019, pp. 6105–6114.

[12] A. Baevski, Y. Zhou, A. Mohamed, and M. Auli, "wav2vec 2.0: A framework for self-supervised learning of speech representations," in Proc. 34th Conf. Neural Information Processing Systems (NeurIPS), 2020.

[13] dima806, "deepfake_vs_real_image_detection" [Online image-classification model]. Hugging Face Model Hub. [Online]. Available: https://huggingface.co/dima806/deepfake_vs_real_image_detection

[14] MelodyMachine, "Deepfake-audio-detection-V2" [Online audio-classification model]. Hugging Face Model Hub. [Online]. Available: https://huggingface.co/MelodyMachine/Deepfake-audio-detection-V2

[15] CardiffNLP, "twitter-roberta-base-sentiment-latest" [Online sentiment-classification model]. Hugging Face Model Hub. [Online]. Available: https://huggingface.co/cardiffnlp/twitter-roberta-base-sentiment-latest

@article{1722274,
      author = {Premdeep Dagar},
      title = {Deepfake Technology in Information Warfare: A Multimodal Detection and Automated Counter-  Propaganda Framework},
      journal = {Iconic Research And Engineering Journals},
      year = {2026},
      volume = {10},
      number = {2},
      pages = {1054-1061},
      issn = {2456-8880},
      url = {https://www.irejournals.com/formatedpaper/1722274.pdf},
      abstract = {Advances in generative deep learning have made photorealistic image, video, and voice synthesis widely accessible, turning "deepfakes" into a practical instrument of information warfare and computational propaganda. Existing tooling largely treats deepfake detection and disinformation-response as separate problems. This paper presents a working prototype that unifies both: a multimodal deepfake detector (image, video, and audio) is combined with a sentiment- and polarization-analysis module over simulated social media reaction, and the two are fused into an automated Information Warfare Threat Score together with drafted counter-propaganda messaging (community notes, headline rebuttals, and verified talking points). Every machine-learning-dependent stage is paired with a deterministic, dependency-light fallback, so the system is fully reproducible without a Graphics Processing Unit (GPU) or model downloads, while transparently reporting which backend (pretrained model vs. heuristic) produced each score. The pipeline is exposed through three interchangeable delivery surfaces, a command-line script, a Representational State Transfer Application Programming Interface (REST API) (FastAPI), and a native desktop Graphical User Interface (GUI) (Tkinter) — that all share the same underlying detection and analysis code. We describe the system architecture, present illustrative results from an end-to-end run in both heuristic and real-model configurations and discuss the limitations and ethical considerations of automating counter-disinformation response.},
      keywords = {Deepfake Detection, Information Warfare, Computational Propaganda, Multimodal Forensics, Sentiment Analysis, Disinformation, Counter-Propaganda, Synthetic Media},
      month = {August},
      doi = {https://doi.org/10.64388/IREV10I2-1722274}
  }