Copyright © 2025 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{185017,
        author = {Dr M K Jayanthi Kannan and N Surya Prakash and K Bala Yaswanth K Siddeswara Reddy and B Rahul A Supreeth},
        title = {Agentic-AI Powered Spam Classifier: An Autonomous SMS Spam Detection Framework with Self-Improving Adaptive Intelligence},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {4},
        pages = {4333-4342},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=185017},
        abstract = {Spam messages are unsolicited messages that can be fraudulent or annoying. The Spam Classifier project is designed to automatically detect and filter spam messages using machine learning and natural language processing techniques. The system converts textual SMS data into numerical feature vectors using techniques like Bag of Words and TF-IDF, and then classifies them as Spam or Ham using Naive Bayes classification. The exponential rise of mobile communication, SMS continues to be a vital medium for personal, commercial, and financial exchanges. However, the increasing prevalence of spam messages poses serious threats to user privacy, security, and trust. Traditional spam detection models, primarily based on static machine learning classifiers, often fail to adapt to evolving spam patterns and adversarial content manipulation. This paper proposes an Agentic AI-powered Spam Classifier, a novel self-improving SMS spam detection framework that integrates large language models (LLMs) with reinforcement learning agents, context-aware embeddings, and adversarial resilience mechanisms. Unlike conventional classifiers, the agentic approach enables autonomous decision-making, dynamic retraining, and continuous adaptation to emerging spam behaviors. The model is trained and tested on real-world SMS datasets to achieve high accuracy, precision, recall, and F1-score. This approach reduces manual intervention in spam detection and provides a scalable solution for real-time spam filtering in communication systems.},
        keywords = {Spam Classification, Natural Language Processing, Text Mining, Naive Bayes, TF-IDF. Agentic AI-powered Spam Classifier, Reinforcement learning agents, context-aware embeddings, Adversarial resilience mechanisms.  Transformer-based deep embeddings (e.g., BERT, RoBERTa), Multi-agent Reinforcement learning, Federated learning layer explainable AI (XAI), Autonomous Agentic Spam detection systems.},
        month = {September},
        }