Large Language Models (LLMs) remain vulnerable to prompt-based attacks such as jailbreaks and prompt injection, highlighting the need for security mechanisms that not only detect malicious intent but also determine whether an attack actually succeeds. In this paper, we introduce Adaptive DeSeTra, a novel security-centric Transformer model designed for deployment that mirrors the real-world attack pipeline through two layers: intent detection via multi-class prompt classification into Benign , Jailbreak and Prompt Injection ; and impact assessment via response-level classification into Compliant or Refusal . Prompt-level intent identification enables low-latency routing to appropriate guardrails, policies, and monitoring controls before a malicious prompt reaches the target Large Language Model (LLM). Conversely, response-level evaluation quantifies whether the adversarial attempt resulted in compliance or refusal, providing an outcome-based measure of attack effectiveness that supports risk assessment and enables more informative security benchmarking than intent detection alone. Experiments demonstrate strong performance across both layers: 99.35% Accuracy/F1/Precision/Recall with 99.85% AUC for prompt classification, and 99.80% for the same metrics with 99.98% AUC for impact assessment, positioning Adaptive DeSeTra as a strong candidate for deployment-oriented LLM security monitoring and evaluation.

The proliferation of IoT networks across various sectors necessitates robust Trust Management mechanisms for secure and reliable operations. This paper proposes a Multi-Attribute Decision Making (MADM)-based approach for trust score calculation in IoT Trust Management. This solution addresses limitations of existing methods by considering multiple attributes and providing a comprehensive evaluation of trustworthiness. The methodology computes a device's trust score by integrating factors such as Cyber Risk, Ease of Access, and Security Level using a weighted sum-based calculation. The Analytical Hierarchy Process (AHP) to determine the factors’ weights is utilized, contributing a novel approach to IoT Trust Management. Furthermore, this approach includes dynamic trust score updates throughout the device's lifetime, accommodating changes in the device's Cyber Risk for accurate trust assessment. A trust score penalization mechanism for devices below a predefined threshold is also introduced, enabling prompt risk mitigation. A simulated assessment, considering varying numbers of IoT devices, evaluates the effectiveness of the proposed methodology. By addressing limitations and introducing innovative components, the proposed MADM-based approach enhances security, reliability, and overall performance of IoT networks. This research advances trust management in IoT and provides valuable insights for developing secure and trustworthy IoT ecosystems.

This paper presents a privacy-enhancing identity management platform designed to address the challenges associated with online identity verification and privacy protection. INCOGNITO offers a comprehensive solution by leveraging concepts such as Qualified Anonymity and cryptographic credentials, along with technologies including blockchain, Tor Network, and software stacks like Idemix. By employing these mechanisms, INCOGNITO aims to enable users to securely acquire and manage their identity attributes, while preserving their privacy and ensuring compliance with both regulatory bodies and Service Providers’ requirements. The platform facilitates the issuance and verification of cryptographic credentials, granting users access to online services based on fine-grained subsets of their identity attributes. Furthermore, the effectiveness and feasibility of the platform are demonstrated through two pilot projects focused on online multimedia content sharing and identifying bots or fake users in online social networks. These pilots showcase the practical applicability of INCOGNITO in solving identity-related challenges while safeguarding user privacy and security.

This paper presents a comprehensive collection of reusable artifacts for addressing security and privacy issues in the context of flow-based programming in Function-as-a-Service (FaaS) environments. With the rapid adoption of FaaS platforms, it becomes important to guarantee the security and privacy of applications. The presented artifacts incorporate a wide variety of nodes and techniques into the popular Node-RED architecture. They intend to improve the security and privacy of applications by addressing critical aspects such as secure data flow management, code authenticity and validation, access control mechanisms, and runtime monitoring and anomaly detection. Using these artifacts, developers can construct more robust and resilient applications in FaaS environments while mitigating potential security and privacy risks.

Digitization, in terms of online services, work environment and other day-to-day procedures, has lead to the wide adoption and use of the respective digital identities. Users utilize their digital personas and their corresponding attributes on a daily basis, in order to gain access to resources and services. This is achieved through the use of numerous identity management schemes, which often suffer from multiple vulnerabilities and are susceptible to threats. This results in the compromise of user privacy and data security. In the recent years, new technologies related to identity management, like the Self-Sovereign Identity (SSI) and eIDAS concepts, are employed to mitigate these issues. This paper presents an architecture that combines state-of-the-art technologies regarding identity management, authentication and secure storage. More specifically, the proposed framework utilizes IOTA-based SSI, the eIDAS framework, FIDO protocol and Trusted Execution Environment (TEE), resulting in a trusted and secure identity management framework. Our solution is thoroughly presented via scenarios, showcasing its robustness and how well it copes in relation to our threat model.

Security for smart industrial systems is prominent due to the proliferation of cyber threats threatening national critical infrastructures. Smart grid comes with intelligent applications that can utilize the bidirectional communication network among its entities. Microgrids are small-scale smart grids that enable machine-to-machine (M2M) communications as they can operate with some degree of independence from the main grid. In addition to protecting critical microgrid applications, an underlying key management scheme is needed to enable secure M2M message transmission and authentication. Existing key management schemes are not adequate due to microgrid special features and requirements. In this article, we propose the Micro sElf-orgaNiSed mAnagement (MENSA), which is the first hybrid key management and authentication scheme that combines public key infrastructure and web-of-trust concepts in microgrids. Our experimental results demonstrate the efficiency of MENSA in terms of scalability and swiftness.