Vulnerability and Attack Injection Guided by Artificial Intelligence, like its base method, consists in injecting realistic vulnerabilities in a component exposed to the Internet and subsequently exploiting such vulnerabilities to launch attacks automatically in order to evaluate existing security mechanisms in the entire system. Vulnerability and Attack Injection Guided by Artificial Intelligence resorts to artificial intelligence to automate the process of defining where to inject vulnerabilities and which attacks to perform as to maximize coverage and reduce execution time when compared to a naïve approach (such as using static code analysis and regex pattern matching). Among the many existing artificial intelligence algorithms, Recurrent Neural Networks (RRNs) and Long Short Term Memory (LSTM) are the most promising techniques for learning the code modifications that are needed to inject a vulnerability (trained using the patches that have been used to fix real-world vulnerabilities) and then to attack these vulnerabilities.