This paper thoroughly explores the importance of fully automated risk assessment and develops an innovative deep learning framework based on advanced artificial intelligence technologies, aimed at enhancing the efficiency and accuracy of risk assessments. This framework utilizes Natural Language Processing (NLP) and Graph Neural Networks (GNN) technologies to automate the processing and analysis of large-scale unstructured and structured data, thereby identifying and evaluating various risks. Through case studies on corporate IPOs and railway transportation system engineering projects, this paper demonstrates the application and effectiveness of the framework in the fields of financial risk assessment and engineering risk assessment. NLP technology allows the framework to deeply understand the complex semantic information contained in text data, while GNN technology enables it to analyze the interaction between experts and the dependency structure within projects, thus providing comprehensive risk assessment results. The deep learning framework developed in this paper not only offers a new automated method for risk assessment but also paves the way for future research and practice in risk management.