A multi-modal framework for detecting fake and cloned websites using holistic web profile
*S. Baghavathi PriyaCorresponding authors_baghavathipriya@ch.amrita.eduDepartment of Computer Science and EngineeringAmrita School of ComputingAmrita Vishwa VidyapeethamChennai, Tamil Nadu, 601103, IndiaView full profile → , D. Pradeep Rajch.en.u4aie22042@ch.students.amrita.eduDepartment of Computer Science and EngineeringAmrita School of ComputingAmrita Vishwa VidyapeethamChennai, Tamil Nadu, 601103, IndiaView full profile → , G. K. Tharun Kaarthikch.en.u4aie22062@ch.students.amrita.eduDepartment of Computer Science and EngineeringAmrita School of ComputingAmrita Vishwa VidyapeethamChennai, Tamil Nadu, 601103, IndiaView full profile → , Venkateswaran RadhakrishnanVenkateswaran.Radhakrishnan@utas.edu.omDepartment of Cyber and Information SecurityCollege of Computing and Information SciencesUniversity of Technology and Applied SciencesSalalah, P.O. Box 608, Dhofar, 211, Sultanate of OmanView full profile → , Syed Arshad Alisyedarshadali1987@gmail.comDepartment of Computer Science & ApplicationsSharda UniversityGreater Noida, Uttar Pradesh, 201310, IndiaView full profile →
* Corresponding author · click or hover a name for details
- Received:
- 01 Dec 2025
- Published Online:
- 30 Sep 2026
- Article type:
- Research Article
- Language:
- EN
- Article no.:
- JIOS-2418
- Pages:
- 3241–3255
Abstract
Loss of digital trust is the primary critical issue that today’s growing and sophisticated phishing campaigns pose against the global digital economy. Such well-orchestrated social engineering attacks effectively imitate legitimate entities so as to trick users into revealing sensitive information. Static blacklist-based and single-feature reliant existing traditional detection approaches are rendered futile against such agile and ephemeral new-age threats. Aegis, an innovative framework, shall be discussed in this paper, which proposes to resolve these issues by conceptualizing a collaborative team of AI cybersecurity experts. This study breaks the rules of typical analysis and introduces a multi-modal approach where AI agents, each one having a unique specialty, analyze a web page from various angles: its URL structure, HTML content, visual design, and brand identity. A real breakthrough is a structured debate mechanism that is set into motion when the initial analyses conflict and the agents reconsider their findings according to the peer arguments in order to resolve ambiguities. Another new feature introduced by this framework is the dynamic real-time verification. The system looks for proof of legitimacy regarding the brand supposedly by an external online search tool to establish an actual basis for comparison. All data is gathered and arranged in LangGraph. Then, Final Adjudicator agent renders an explainable verdict on whether the site is legitimate. This multi-agent-driven investigative process makes Aegis more robust and human-like in zero-day phishing threat detection. This study initiates the future generation of capable active intelligent defence systems able to manoeuvre within an ever-evolving cyber deception scenario.
Keywords
Subject Classifications
References
[1] M. Vijayalakshmi, S. Mercy Shalinie, M. H. Yang, and U. R. Meenakshi, “Web phishing detection techniques: A survey on the state-of-the-art, taxonomy and future directions,” IET Networks, vol. 9, no. 5, pp. 235–246 (2020).
[2] R. S. Rao and S. T. Ali, “PhishShield: A desktop application to detect phishing webpages through heuristic approach,” Procedia Computer Science, vol. 54, pp. 147–156 (2015).
[3] M. G. Alkhozae and O. A. Batarfi, “Phishing websites detection based on phishing characteristics in the webpage source code,” International Journal of Information and Communication Technology Research, vol. 1, no. 6 (2011).
[4] A. K. Jain and B. B. Gupta, “Phishing detection: Analysis of visual similarity based approaches,” Security and Communication Networks, vol. 2017, Art. no. 5421046 (2017).
[5] J. Mao, W. Tian, P. Li, T. Wei, and Z. Liang, “Phishing-Alarm: Robust and efficient phishing detection via page component similarity,” IEEE Access, vol. 5, pp. 17020–17030 (2017).
[6] W. Zhang, H. Lu, B. Xu, and H. Yang, “Web phishing detection based on page spatial layout similarity,” Informatica, vol. 37, no. 3 (2013).
[7] A. Aljofey, Q. Jiang, A. Rasool, H. Chen, W. Liu, Q. Qu, and Y. Wang, “An effective detection approach for phishing websites using URL and HTML features,” Scientific Reports, vol. 12, no. 1, Art. no. 8842 (2022).
[8] A. Subasi and E. Kremic, “Comparison of AdaBoost with MultiBoosting for phishing website detection,” Procedia Computer Science, vol. 168, pp. 272–278 (2020).
[9] Y. Li, Z. Yang, X. Chen, H. Yuan, and W. Liu, “A stacking model using URL and HTML features for phishing webpage detection,” Future Generation Computer Systems, vol. 94, pp. 27-39 (2019).
[10] C. Opara, B. Wei, and Y. Chen, “HTMLPhish: Enabling phishing web page detection by applying deep learning techniques on HTML analysis,” in Proc. Int. Joint Conf. Neural Networks (IJCNN), pp. 1-8 (2020).
[11] J. Feng, L. Zou, O. Ye, and J. Han, “Web2Vec: Phishing webpage detection method based on multidimensional features driven by deep learning,” IEEE Access, vol. 8, pp. 221214-221224 (2020).
[12] Y. Wei, M. Nakayama, and Y. Sekiya, “Enhancing generalization in phishing URL detection via a fine-tuned BERT-based multimodal approach,” IEEE Access, vol. 13, pp. 131197-131216 (2025).
[13] Z. Yao, I. Florescu, and C. Lee, “Control in stochastic environment with delays: A model-based reinforcement learning approach,” arXiv preprint arXiv:2402.00313 (2024).
[14] C. L. Tan, K. L. Chiew, K. Wong, and S. N. Sze, “PhishWHO: Phishing webpage detection via identity keywords extraction and target domain name finder,” Decision Support Systems, vol. 88, pp. 18-27 (2016).
[15] G. Liu, B. Qiu, and L. Wenyin, “Automatic detection of phishing target from phishing webpage,” in Proc. 20th Int. Conf. Pattern Recognition (ICPR), pp. 4153-4156 (2010).
[16] R. S. Rao, T. Vaishnavi, and A. R. Pais, “PhishDump: A multi-model ensemble based technique for the detection of phishing sites in mobile devices,” Pervasive and Mobile Computing, vol. 60, Art. no. 101084 (2019).
[17] J. H. Yoon, S. J. Buu, and H. J. Kim, “Reinforced disentangled HTML representation learning with hard-sample mining for phishing webpage detection,” Electronics, vol. 14, no. 6, Art. no. 1080 (2025).
[18] M. Wang, L. Song, L. Li, Y. Zhu, and J. Li, “Phishing webpage detection based on global and local visual similarity,” Expert Systems with Applications, vol. 252, Art. no. 124120 (2024).
[19] W. Li, S. Manickam, Y. W. Chong, and S. Karuppayah, “PhishDebate: An LLM-based multi-agent framework for phishing website detection,” arXiv preprint arXiv:2506.15656 (2025).
[20] W. Li, S. Manickam, Y. W. Chong, and S. Karuppayah, “PhishIntentionLLM: Uncovering phishing website intentions through multi-agent retrieval-augmented generation,” arXiv preprint arXiv:2507.15419 (2025).
[21] T. Cao, C. Huang, Y. Li, W. Huilin, A. He, N. Oo, and B. Hooi, “PhishAgent: A robust multimodal agent for phishing webpage detection,” in Proc. AAAI Conf. Artificial Intelligence, vol. 39, no. 27, pp. 27869-27877 (Apr. 2025).
[22] V. Sellam, M. R. Kumar, V. S. Kumar, S. M. Kumar, and G. T. Kaarthik, “Novel lightweight encryption scheme based on transformation matrix and ChaCha20 algorithm for secure snap transmission,” in Proc. Int. Conf. Intelligent Systems and Advanced Applications (ICISAA), pp. 1-6 (2024).
[23] M. Ranjith Kumar, V. Sudeesh Kumar, G. K. Tharun Kaarthik, V. Chatiyode, J. Srinivasan Anusha, and P. Revathi, “Snowfall forecasting with enhanced accuracy: Leveraging multi-class SVM for meteorological predictions,” in Proc. Int. Conf. Computational Intelligence in Data Science, Cham, Switzerland: Springer Nature, pp. 312-328 (2024).
[24] V. Addanki, S. Durgapu, K. Dorasanaiah, and S. Abhishek, “Safeguarding SMS: A dynamic duo approach to tackle spam using LDA and QDA,” in Proc. Innovations in Power and Advanced Computing Technologies (i-PACT), pp. 1-6 (2023).
[25] M. Asmitha and C. R. Kavitha, “Exploration of automatic spam/ham message classifier using NLP,” in Proc. IEEE 9th Int. Conf. Convergence in Technology (I2CT), pp. 1-7 (2024).
[26] R. R. Patil, G. Kaur, H. Jain, A. Tiwari, S. Joshi, K. Rao, and A. Sharma, “Machine learning approach for phishing website detection: A literature survey,” Journal of Discrete Mathematical Sciences and Cryptography, vol. 25, no. 3, pp. 817-827 (Apr. 2022), doi: 10.1080/09720529.2021.2016224.
