As technology has improved cyber criminals have developed newer and more sophisticated methods of attack. As a result cybersecurity professionals have needed to adapt and improve their own methods of defending against cyber threats. One technology that is increasingly being leveraged against such threats is machine learning. Machine learning is an aspect of artificial intelligence (AI) and computer science that uses data and algorithms to build models of underlying patterns allowing for the prediction of future data and classifying current data. In the context of cybersecurity machine learning can be used in a variety of ways including monitoring activity within a network in order to detect malicious or abnormal activity, monitoring background activity of individual computers in order to detect malware, as well as other uses. By collecting data and building a model of normal patterns, machine learning allows cybersecurity professionals to automate the process of monitoring systems in real time in order to immediately detect abnormal activity and rapidly respond to threats and breaches. Its use however is not limited to cybersecurity professionals but also by cyber criminals. Cyber attacks are now being executed using machine learning as well. Increasingly sophisticated machine learning models are being used to execute attacks in new ways. Some ways machine learning is being used in cyber attacks include using it to evade web filters, bypass CAPTCHA checks, and creating targeted phishing emails and messages. As the technology has evolved newer methods are being used both in offense and defense and recent trends indicate that it will continue to be an important subject in regards to the future of the cybersecurity field. In this paper I will endeavor to create a comprehensive review and explanation of how this technology is being used both by cybersecurity professionals and cyber criminals, its strengths and weaknesses in regards to current practices and future trends.