The field of software engineering is currently trending toward the high demand for global software development. The idea of employing a software engineering specialist from anywhere in the world with a variety of skills and expertise to meet needs and at an affordable price is the main driver behind the fastest-growing global software development approach. On the other hand, it can be difficult to integrate distributed teams with a company's resources and tools. Therefore, a precise assessment of the risks associated with the software project is required, along with early risk prediction. This comprehensive literature review provides an overview of various risk prediction models used in global software development. This literature review discusses 12 studies that use Many models and techniques such as machine learning, neural networks, mathematics, algorithms, similarity analysis, and frameworks that try to predict software failures and risks. In addition, this research goes into depth and provides suggestions for improving machine learning models and frameworks for future studies