Social engineering (SE) attacks are a growing concern for organizations that rely on technology to protect sensitive data. Identifying and preventing these attacks can be challenging, as they frequently rely on manipulating human behavior rather than exploiting technical vulnerabilities. Although various studies have explored SE attacks and their defense mechanisms, there remains a gap in the literature concerning the holistic and layered classification of these threats and countermeasures. To address this, we conducted a comprehensive literature survey to understand existing taxonomies and subsequently identified areas that required a more structured and exhaustive categorization. Based on the survey results, we propose a comprehensive taxonomy of SE attacks, classifying them based on three levels: environment, approaches, and mediums. Additionally, we present a taxonomy of social engineering countermeasures, encompassing both technical and non-technical solutions. The proposed taxonomies serve as a foundation for future research and offer organizations a valuable framework for developing effective strategies to detect, prevent, and respond to social engineering incidents.

Nowadays, web application security is one of the relevant issues in the IT security domain due to the continued growth in the number of web-related attacks. As a result, attacks, with various and varied motivations, have developed and become increasingly sophisticated. They mainly target data related to economic activities. Thus, they cause significant damage to the overall functioning of information systems. To address the various threats, several robust taxonomies exist in the literature. Each taxonomy and classification has advantages and limitations. We first define the different threat classifications related to the context of Web applications. The objective of this analysis is to provide a synthesis of the advantages and disadvantages of each classification. The current work analyses different taxonomies for web applications threats, in order to propose our proper taxonomy. The proposed taxonomy takes advantage of the benefits of existing taxonomies and provides an integrated approach for classifying both client-side and server-side attacks. The finding will help researchers to find a clear and detailed taxonomy of the different threats related to web applications.

OpenID Connect (OIDC) is one of the most widely used delegated authentication protocols in web and mobile applications providing a single sign-on experience. It allows third-party applications, called Relying Parties (RP), to securely request and receive information about authenticated sessions and end-users from an identity provider. The OIDC specification defines several parameters, including the client_id, client_secret, authorization code, access token, id token, state, and redirect_uri, as keys to the protocol operation, with significant security and privacy implications. Therefore, securing these parameters is critical to prevent attackers from impersonating legitimate entities, gaining unauthorized access, having complete control over users’ accounts, and/or violating their privacy. To enhance OIDC security and preserve its users’ privacy, we propose a novel model for OIDC based on the Ethereum Blockchain and the non-fungible token (ERC721) standard. To prove the robustness and safety of the proposed system, we perform a detailed security analysis formally using the most widely accepted protocols security verification tools, AVISPA and Scyther, and informally by discussing various attacks. The analysis results show that the proposed system is resilient against well-known attacks. Furthermore, we evaluate the cost and performance of the proposed solution, confirming its affordability and assuring that our approach does not impact the user experience and performance of existing OIDC-based systems. Finally, we conduct a security and privacy comparative analysis with similar existing systems, proving the superiority and efficiency of our proposed Blockchain-based OIDC system.