Malware is a significant and growing threat to the security of computer systems in various domains. Malware detection has become a critical aspect of safeguarding personal and organisational data. As the malware continues to rise, traditional detection methods cannot cope with the sophistication of these attacks. These conventional methods often fail to recognise new or polymorphic malware, compromising the effectiveness of providing thorough protection. Dynamic and hybrid models have proven superior accuracy and adaptability, especially when trained on large datasets. This paper discusses several techniques for malware detection and the promising potential of the machine and deep learning models. It also makes a comparative analysis of traditional methods and techniques based on dynamic and hybrid approaches.