Today, artificial intelligence (AI) is a very popular topic of discussion. Many well-known contributing scientists and future theorists have been weighing in and reflecting on the potential threats to humans by the more advanced superintelligence AI systems. However, AI systems are based on computer programming languages. Therefore, just as in any other computer programming, AI is very vulnerable to cyber threats and thereby very exploitable. AI will have many threats in the future in the form of insecure coding which has created various vulnerabilities. There are two types of significant AI in existence today. They are narrow (weak) AI and general AI. The technology of superintelligence AI or more advanced AI systems are still emerging. Narrow AI embedded robotic systems such as a single task mission robots; The LEGO Mindstorms EV3 robotic system has been tested in this quantitative-experimental dissertation study for the existence of any type of implemented security architecture. The LEGO Mindstorms EV3 robotic system has a very narrow focused mission. Thereby, it has not been determined if it is using verified program system functions or a computer security architecture during its performance. Therefore, narrow AI programmed systems which have been written in the popular programming codes are susceptible to cyber-attacks as any other implemented computer software application. The dissertation research study as an experiment has scanned the LEGO Mindstorms EV3 robotic system for various vulnerabilities to attempt penetration testing in order to prove a better developed security architecture design is needed for the future performance of other emerging AI robotic control systems, as they come online in the future. The dissertation research study will conclude with recommended mitigating factors such as a proposed cybersecurity architecture which has been designed to prevent the type of narrow AI cyberattacks from targeting such systems in the future and beyond.