The interface of quantum computing with classical machine learning through Quantum Machine Learning (QML) creates a new research opportunity for better computational techniques. The paper conducts a systematic literature investigation to understand the fundamental association between QML security elements and aspects. Multiple security assessments of QML models receive evaluation in this review by exploring both quantum architectural weaknesses followed by surveys of suggested security measures. Our analysis introduces security vulnerabilities that do not exist in conventional platforms. The defensive capabilities of QML models increase when trained with adversarial methods and when quantum noise is utilized together with differential privacy techniques. Further analysis will be crucial since it provides fundamental protection for QML systems as they are implemented across practical applications. Every new research project and practical security application in QML security benefits from starting their investigations with this work.