The paper concludes that the convergence of trustworthy AI, robust multimedia analysis, and privacy-preserving architectures is essential for responsible AI adoption in regulated domains, and highlights open challenges in cross-document synthesis, multimodal robustness, and scalability for future research.
This study presents a detailed, actionable approach to constructing secure, privacy-focused RAG systems and culminates in the Integrated Privacy-Preserving RAG Framework (IPRAG), a five-tier architecture supported by a three-phase deployment protocol.
Firoz Mohammed Ozman· International Journal of Fro...· 0 citations
Large language models (LLMs) are increasingly embedded as core components of data-centric systems, supporting analytical decision making, and automated reasoning over large-scale, heterogeneous datasets. Yet their deployment in open-world environments raises fundamental challenges to security and trustworthiness: LLMs...
Lu Lin, Jinghui Chen, Ting Wang et al.· Proceedings of the 32nd ACM...· 0 citations
CTRAG is presented, a novel Retrieval-Augmented Generation pipeline designed for automated compliance checking that employs advanced strategies, including adaptive chunking, dynamic retrieval configurations, and in-context learning, to improve the precision and relevance of compliance assessments.
Muhammad Roman, Karen Rafferty, Barry Devereux· 0 citations
Recommendations for the design and deployment of secure LLM-enabled XR systems, emphasizing strategies such as security defense in depth, privacy-aware data practices, layered technical safeguards, continuous monitoring, and governance mechanisms that align security controls with the specific risks of immersive and int...
This paper presents a graph-based framework that transforms this static corpus of O-RAN security threats into a single, queryable database and integrates a hybrid data extraction pipeline that couples deterministic parsing of structured specifications with Large Language Model (LLM)-assisted extraction for evolving spe...
Corban Villa, Michele Guerra, Syed Khandker et al.· 0 citations
Cloud misconfiguration remains a leading cause of security incidents, yet whether LLMs and SLMs can generate security-compliant Infrastructure-as-Code is an open question. We benchmark seven models, three closed LLMs (Claude Opus 4, GPT-5.4, Gemini 2.5 Pro) and four open SLMs (Qwen2.5-Coder-14B, WizardCoder-33B, CodeLl...
Francis Luis Santos Vargas, R. Mansilha, Diego Kreutz· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
