In recent years, various companies from Target to Yahoo have been hacked for personal user information. Hacking is an increasingly prevalent issue that costs both companies and consumers billions of dollars each year. In response, cyber security firms continually update their mechanisms for securing user data against malicious software. The methodology used to prevent such attacks begins with structuring software with threat mitigation built in, deciding which attacks are necessary to prevent, and predicting and covering the vulnerabilities of the software. One such vulnerability is evident in XML files, which at their most basic are used to store data within HTML code and at their most complex can be used to carry out attacks such as the 2007 Dridex Trojan that utilized the versatility of the XML file structure to steal users’ bank data. This versatility makes XML attacks especially difficult to defend against, and this project seeks to develop a methodology to mitigate such attacks. This will be done by extracting the relevant features from a file, running statistical analysis on these features, then creating a model from the data of a normal XML file which aids in deciding which files are malicious by recognizing deviation from the model. Due to the number of attacks that can be carried out with XML such as coercive parsing and parameter tampering and XML’s growing role in developing web applications, securing networks against malicious XML files will aid in maintaining information security and prevent costly attacks on companies and users.