privacy threat modeling

Attack scenario descriptions in cybersecurity automation lack formal semantic foundations, hindering systematic analysis and automation. Method: This paper proposes an abstract, formal model based on UML class diagrams, enabling the first unified modeling of attack context and attack scenarios. The model supports structured input, automated processing, and cross-process reuse, directly facilitating two core tasks: attack analysis and automated attack script generation. Contribution/Results: Evaluated on real-world attack analysis and cybersecurity training script generation, the model demonstrates strong feasibility and effectiveness. It fills a critical gap in formal attack scenario modeling and establishes a scalable, verifiable semantic foundation for security process automation—enhancing interoperability, reproducibility, and formal reasoning in cyber defense systems.

Digital forensics is critical in criminal investigations, yet evidence acquisition, processing, and storage frequently entail significant privacy risks; existing research lacks systematic threat identification and modeling, leading to regulatory noncompliance and rights violations. This paper introduces the SPADA methodology—previously unapplied in digital forensics—to establish the first privacy threat model that rigorously integrates legal compliance and ethical requirements. Through legal compliance analysis, cross-jurisdictional comparative assessment, and structured threat elicitation, we identify 298 domain-specific threats alongside several cross-cutting, generic threats. The resulting model enables threat classification, provenance tracing, and automated compliance evaluation. It has been operationalized in forensic tool design and policy development, thereby bridging a critical gap between privacy protection theory and judicial practice.

Existing privacy threat modeling approaches adopt a single-user-centric paradigm, rendering them inadequate for capturing Interdependent Privacy (IDP)—a phenomenon wherein one user’s actions inadvertently compromise the privacy of others, including non-users. This challenge is exacerbated in third-party application ecosystems due to permission misuse, unregulated API invocations, and cross-entity behavioral coupling. To address this gap, we propose IDPA, the first threat modeling method explicitly designed for IDP. IDPA extends the LINDDUN framework by integrating permission flow analysis, API call graph modeling, and multi-stakeholder behavioral modeling—thereby shifting from a monolithic to a relational, system-level perspective. An empirical study conducted within the WeChat ecosystem demonstrates that IDPA systematically identifies six novel IDP threat categories—such as passive exposure of non-user data and authorization chain leakage—that are entirely missed by conventional methods. Results show a substantial improvement in privacy risk detection capability, establishing a new paradigm for privacy governance in platformized digital environments.

Smart home systems suffer from inadequate protection of user and device identity data privacy, coupled with fragmented threat analysis and risk management practices. Method: This paper proposes the first privacy engineering framework that centrally incorporates device identity privacy. It integrates Data Flow Diagrams (DFDs) with LINDDUN PRO for end-to-end threat modeling, and combines Privacy Impact Assessments (PIAs) with Privacy-Enhancing Technologies (PETs) to systematically identify, quantitatively assess, and prioritize privacy risks. Contributions: (1) It establishes— for the first time—the critical role of device identity privacy within smart home privacy engineering; (2) it introduces an actionable risk prioritization matrix and standardized mitigation strategies; and (3) it delivers implementable privacy governance pathways for device manufacturers, cloud service providers, end users, and regulatory authorities.

论文提出ThreatModeling-LLM框架，利用大语言模型自动化银行业系统的威胁建模，解决传统方法效率低、易出错的问题，通过数据集创建、提示工程和模型微调三阶段实现。

This study addresses the security risks posed by AI agents with offensive cyber capabilities that may breach sandbox boundaries in evaluation environments. It systematically identifies five categories of boundary vulnerabilities—multi-step attacks, objective conflicts, supply chain leaks, persistence mechanisms, and automated execution speed—and conducts a case analysis grounded in the 2026 Hugging Face/OpenAI incident. The work introduces the first taxonomy of AI boundary vulnerabilities specifically tailored to evaluation settings and proposes an integrated defense framework combining isolation, privilege separation, behavioral provenance tracking, and defensive response interfaces. By jointly considering misuse risks and capability assessment, this research establishes clear security priorities for high-risk AI evaluations, offering both theoretical foundations and practical guidance for developing trustworthy evaluation environments that balance testing efficacy with risk containment.

This work addresses the evolving privacy threats faced by AI systems throughout their lifecycle, which encompass both traditional risks and emerging model-centric attacks—such as membership inference and model inversion—that existing frameworks like LINDDUN fail to comprehensively cover. To bridge this gap, the authors propose a hybrid modeling paradigm that integrates LINDDUN with a knowledge base of AI-specific attacks. By leveraging metadata from data flow diagrams, semantic retrieval via a vector database, and data flow–aware prompt engineering, the approach guides large language models to perform lifecycle-aware, knowledge-driven, and interpretable privacy threat identification and classification. Empirical evaluation demonstrates that the method not only covers all classical privacy threat categories but also effectively detects model-centric attacks, yielding consistent and knowledge-grounded outputs across different large language models.

This study addresses the lack of objective validation criteria in existing threat modeling approaches, which often rely on expert judgment and are thus prone to omissions or inconsistencies. To overcome this limitation, the authors propose a quantifiable and reproducible evaluation methodology based on benchmark applications with known vulnerabilities—specifically AzureGoat and VulnBank. Using only architectural diagrams, data flow diagrams, and their textual descriptions as input, the approach evaluates the vulnerability coverage of ThreMoLIA, an LLM-assisted threat modeling system, against Microsoft Threat Modeling Tool. Experimental results demonstrate that ThreMoLIA achieves consistently higher vulnerability coverage across both benchmark applications. This work represents the first effort to employ real-world vulnerable applications as a validation benchmark for threat modeling, effectively mitigating the shortcomings inherent in traditional expert-based assessments.

Current technology vendors commonly lack granular security disclosures, issuing only vague “security statements,” thereby impeding visibility and assurance across the software supply chain. Method: This paper introduces the Public Threat Model (PTM) paradigm—a systematic approach to enhance cross-organizational security transparency and collaborative defense. We design a PTM construction framework applicable to both open-source and commercial components, incorporating threat modeling, sensitive information anonymization, versioned updates, and regulatory compliance alignment. We further provide actionable publication guidelines and review procedures. Contribution/Results: Our work enables the transformation of proprietary internal threat models into standardized, verifiable, and integrable public assets. This empowers procuring organizations to rigorously assess security posture and allows vendors to differentiate through demonstrable security capabilities—establishing a novel, practice-grounded paradigm for building trustworthy software supply chains.

This work addresses the inefficiency and manual effort inherent in threat modeling for cyber-physical systems (CPS) by proposing SMSI, an automated framework that integrates neural and symbolic methods. SMSI enables the first-ever automatic mapping from SysML architectural models to NIST SP 800-53 security controls through an end-to-end reasoning chain linking CVE, MITRE ATT&CK, and NIST frameworks. The approach combines a deterministic parser, a fine-tuned SecureBERT+ classifier, a dense retrieval encoder, and the Gemma-26B large language model for zero-shot inference to generate prioritized security control recommendations. Evaluation on a medical IoT gateway case study demonstrates that SecureBERT achieves superior performance in recommending NIST controls from ATT&CK tactics, confirming that dense embeddings effectively support automated security modeling for CPS.