• Adhrija T S
• Dr.Sudheer S Marar

Machine learning (ML) technologies have revolutionized the generally adopted cybersecurity applications like malware detection, intrusion prevention, and biometric authentication. But in doing so, growing reliance on ML models only put them in the harm's way of adversarial machine learning (AML) attacks in which imperceptible manipulations of input data can deceive the model, bringing along with it extremely severe consequences such as unapproved access and system failure. The stealthy nature of these complex and constantly changing threats makes them difficult to neutralize with conventional cybersecurity measures. This research introduces state-of-the-art defense approaches, including adversarial training, curriculum learning, and adaptive mechanisms, which synergistically enhance the robustness and reliability of ML models. With systematic resilience enhancements and dynamic adaptability to ever-changing adversarial techniques, such approaches ensure integrity in ML-based cybersecurity frameworks against emerging threats

Adversarial Machine Learning, Cybersecurity, Adversarial Attacks, Adversarial Training, Curriculum Learning, AI Security

"FORTIFYING AI: COMBATING ADVERSARIAL THREAT IN CYBERSECURITY", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 3, page no.b471-b476, March-2025, Available :http://www.jetir.org/papers/JETIR2503153.pdf

"FORTIFYING AI: COMBATING ADVERSARIAL THREAT IN CYBERSECURITY", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 3, page no. ppb471-b476, March-2025, Available at : http://www.jetir.org/papers/JETIR2503153.pdf