Automated Security Testing for Identity Management of Large-scale Digital Infrastructures

Ensuring the security of an organization's digital assets against cyber threats is critical in today's technology-driven world. Regular security testing is one of the measures that can help assess the effectiveness of security controls, identify vulnerabilities, and strengthen the overall cybersecurity posture. Identity Management (IdM) protocols such as Security Assertion Markup Language 2.0, OpenID Connect, and OAuth 2.0 play a crucial role in protecting against identity theft, fraud, and security breaches. Also, following the Best Current Practices introduced by the standards to enhance the security of IdM protocols is essential to minimize the risk of unauthorized access, data breaches, and other security threats and to maintain compliance with regulatory requirements, and build trust with users and stakeholders. However, deploying these protocols can be challenging due to the complexity in designing, developing and implementing cryptographic mechanisms. The implementation of IdM protocols encounters three significant obstacles: fragmented security information, rapidly evolving threat environment, and the need for a controlled testing environment. Security testers must stay up-to-date with emerging threats and establish an appropriate testing infrastructure to guarantee the security and robustness of IdM implementations, while also minimizing the possibility of security incidents that could adversely affect operations. Automated security testing plays a crucial role in addressing security concerns, particularly as the intricate functional aspects of IdM solutions contribute to their complexity. It is essential to prioritize automation to bridge the cybersecurity skills gap among IT professionals. In this thesis, we propose Micro-Id-Gym (MIG), a framework that offers (i) an easy way to configure and reproduce the IdM production environment in a sandbox, allowing hands-on experiences with potentially impactful security tests that may winder availability of services and (ii) automatic security testing of IdM implementations together with suggestions for mitigations to avoid identified vulnerabilities. MIG provides a set of security testing tools for creating, executing, and analyzing security test cases through MIG-L, a declarative test specification language. We have evaluated the effectiveness of MIG by conducting experiments to assess the accuracy in supporting detection of relevant vulnerabilities in the implementation of IdM protocols. We utilized MIG to conduct security analyses across various corporate scenarios and projects, identifying vulnerabilities and responsibly disclosing them through bug bounty programs. Our findings were recognized by the providers, who awarded us both monetary compensation and public recognition. Overall, MIG can help organizations establish a robust and agile security testing strategy, supported by suitable infrastructure and testing procedures, that can ensure the security and resilience of their IdM implementations.