Copyright © 2026 Authors retain the copyright of this article. This article 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.

@article{208745,
        author = {MAHANKALI RAVISANKAR and Dr. B. Anuradha},
        title = {Detection Of AI-Generated Phishing Emails Using Hybrid Attention-Based NLP Models},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {4},
        pages = {2964-2973},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=208745},
        abstract = {In recent years, artificial intelligence and natural language processing have transformed the field of cybersecurity, particularly in the detection of phishing emails. This research focuses on detecting AI-generated and traditional phishing emails that closely resemble legitimate messages and are difficult for conventional filtering systems to identify. The proposed system combines advanced text preprocessing with TF-IDF and RoBERTa to extract both statistical and contextual features from email content. These features are fused with URL-based information and balanced using SMOTE-Tomek to improve classification performance. A BiLSTM model is employed for deep feature learning, while XGBoost is used for final email classification and performance evaluation. The model is further optimized using Optuna, validated through Stratified K-Fold cross-validation, and interpreted using SHAP explainability. The proposed approach achieved 99.29% accuracy, 99.23% precision, 99.36% recall, 99.30% F1-score, and 99.96% ROC-AUC, demonstrating its effectiveness in identifying phishing emails. However, the system still faces challenges such as false predictions, computational cost, dataset limitations, and the need for continuous adaptation to evolving AIgenerated phishing attacks.},
        keywords = {AI Generated Phishing Emails, Phishing Detection, Natural Language Processing (NLP),TFIDF, RoBERTa, BiL STM, XGBoost, SMOTE-Tomek, Optuna, SHAP Explainability.},
        month = {September},
        }