Towards Architectural Pen Test Case Generation and Attack Surface Analysis to Support Secure Design

In today's interconnected world, software systems have become indispensable components of complex frameworks, such as cyber-physical systems, e-commerce platforms, and healthcare information systems. This widespread integration highlights the importance of security more than ever, as these systems often function in critical environments where vulnerabilities can lead to significant destructive consequences. Pen-etration testing is a key method for identifying vulnerabilities but is often conducted after deployment, making remediation costly and time-consuming. On the other hand, back to the early phases of the Software Development Life Cycle (SDLC), software architects often lack the security expertise and feedback mechanisms needed to make informed design decisions, leading to vulnerabilities that remain undetected until later stages. In this paper, to address these challenges, we propose a research plan to integrate security considerations into the design phase. Our approach involves generating architecture-based penetration test cases, evaluating the attack surface of alternative architectures by using the generated test cases, and supporting software architects in making secure design decisions afterward. We plan to leverage threat modeling and Large Language Models (LLMs) to fulfill our target ambitions. To validate the applicability and effectiveness of our approach, we aim to conduct case studies in areas such as autonomous vehicles (AVs), which present significant security challenges.