The large branches of Machine Learning represent an immense support for the detection of malicious websites, they can predict whether a URL is malicious or benign, leaving aside the cyber attacks that can generate for network users who are unaware of them. The objective of the research was to know the state of the art about Neural Networks and their impact for the Detection of malicious Websites in network users. For this purpose, a systematic literature review (SLR) was conducted from 2017 to 2021. The search identified 561 963 papers from different sources such as Taylor & Francis Online, IEEE Xplore, ARDI, ScienceDirect, Wiley Online Library, ACM Digital Library and Microsoft Academic. Of the papers only 82 were considered based on exclusion criteria formulated by the author. As a result of the SLR, studies focused on machine learning (ML), where it recommends the use of algorithms to have a better and efficient prediction of malicious websites. For the researchers, this review presents a mapping of the findings on the most used machine learning techniques for malicious website detection, which are essential for a study because they increase the accuracy of an algorithm. It also shows the main machine learning methodologies that are used in the research papers.