As artificial Intelligence (AI) continues to expand exponentially, particularly with the emergence of generative pre-trained transformers (GPT) based on a transformer’s architecture, which has revolutionized data processing and enabled significant improvements in various applications. This document seeks to investigate the security vulnerabilities detection in the source code using a range of large language models (LLM). Our primary objective is to evaluate the effectiveness of Static Application Security Testing (SAST) by applying various techniques such as prompt persona, structure outputs and zero-shot. To the selection of the LLMs (CodeLlama 7B, DeepSeek coder 7B, Gemini 1.5 Flash, Gemini 2.0 Flash, Mistral 7b Instruct, Phi 3 8b Mini 128K instruct, Qwen 2.5 coder, StartCoder 2 7B) with comparison and combination with Find Security Bugs. The evaluation method will involve using a selected dataset containing vulnerabilities, and the results to provide insights for different scenarios according to the software criticality (Business critical, non-critical, minimum effort, best effort) In detail, the main objectives of this study are to investigate if large language models outperform or exceed the capabilities of traditional static analysis tools, if the combining LLMs with Static Application Security Testing (SAST) tools lead to an improvement and the possibility that local machine learning models on a normal computer produce reliable results. Summarizing the most important conclusions of the research, it can be said that while it is true that the results have improved depending on the size of the LLM for business-critical software, the best results have been obtained by SAST analysis. This differs in “Non-Critical,” “Best Effort,” and “Minimum Effort” scenarios, where the combination of LLM (Gemini) + SAST has obtained better results.

In today’s rapidly evolving digital landscape, web application security has become paramount as organizations face increasingly sophisticated cyber threats. This work presents a comprehensive methodology for implementing robust security measures in modern web applications and the proof of the Methodology applied to Vue.js, Spring Boot, and MySQL architecture. The proposed approach addresses critical security challenges through a multi-layered framework that encompasses essential security dimensions including multi-factor authentication, fine-grained authorization controls, sophisticated session management, data confidentiality and integrity protection, secure logging mechanisms, comprehensive error handling, high availability strategies, advanced input validation, and security headers implementation. Significant contributions are made to the field of web application security. First, a detailed catalogue of security requirements specifically tailored to protect web applications against contemporary threats, backed by rigorous analysis and industry best practices. Second, the methodology is validated through a carefully designed proof-of-concept implementation in a controlled environment, demonstrating the practical effectiveness of the security measures. The validation process employs cutting-edge static and dynamic analysis tools for comprehensive dependency validation and vulnerability detection, ensuring robust security coverage. The validation results confirm the prevention and avoidance of security vulnerabilities of the methodology. A key innovation of this work is the seamless integration of DevSecOps practices throughout the secure Software Development Life Cycle (SSDLC), creating a security-first mindset from initial design to deployment. By combining proactive secure coding practices with defensive security approaches, a framework is established that not only strengthens application security but also fosters a culture of security awareness within development teams. This hybrid approach ensures that security considerations are woven into every aspect of the development process, rather than being treated as an afterthought.