AI has emerged as a new medium for the dissemination of animosity via various social media platforms. This study investigates how advanced AI technologies are assisting in the prediction and analysis of online hate speech trends. Using linear & multiple regression, KNN & clustering algorithms, this study seeks to determine the relationship between hate speeches and other variables. Machine Learning algorithms, an important subfield of Artificial Intelligence, are an efficient method for discovering object relationships. This paper also discusses internet-based hate speeches about artificial intelligence. Importantly, the researchers investigate through a comprehensive literature review that combating hate speeches requires first monitoring and analyzing their dynamics to comprehend their nature. Since the spread of malicious communication can be an early indicator of violence, particularly heinous crimes, reducing hate speeches could help reduce their effects. Consequently, this paper examines the targets of hatred on the Internet and proposes a framework for detecting problems using AI and ML technologies. By evaluating alternative parameters and non-parameters, algorithms can be used to detect and prevent internet-based hate speeches, according to the study’s conclusion. To establish a secure and wholesome environment. This paper also examines the international conventions which include two legal regimes CRED and ICCPR which are working on hate speech issues around the globe acting as a guideline for the nations.