Cybersecurity & Exploits

Global: DoS Attacks and Defense Technologies in Blockchain Systems: A Hierarchical AnalysisA research paper authored by Chunyi Zhang, Fengjiao Dou, and Xiaoqi Li was first submitted to arXiv on July…

Global: Audio Narrative Attacks Bypass Safety Controls in Large Audio-Language ModelsResearchers Ye Yu, Haibo Jin, Yaoning Yu, Jun Zhuang, and Haohan Wang reported a novel security vulnerability affecting large audio‑language…

Global: Study Finds Over 22% of Analyzed Android Apps Vulnerable to TLS Man-in-the-Middle AttacksA new study released on arXiv reports that 8,374 out of 37,349 Android applications examined contain TLS…

Global: AEGIS Leverages Large Language Models to Automate Attack Path Creation for Cyber‑Defense ExercisesResearchers Ivan K. Tung, Yu Xiang Shi, Alex Chien, Wenkai Liu, and Lawrence Zheng introduced a system…

Global: Researchers Introduce FaceDefense to Counter Diffusion-Based Face SwappingOn January 30, 2026, Yilong Huang and Songze Li submitted a paper to arXiv describing FaceDefense, a proactive defense framework designed to protect…

Global: Researchers Propose AlienLM to Shield LLM API InteractionsOn January 30, 2026, researchers Jaehee Kim and Pilsung Kang submitted a preprint to arXiv describing a novel privacy layer called AlienLM…

Global: New Benchmark Highlights LLM Web Agents’ Susceptibility to Malicious URLsResearchers led by Dezhang Kong released MalURLBench on January 26, 2026, a comprehensive benchmark designed to assess large‑language‑model (LLM) web…

Global: Study Reviews Methods and Tools for System Penetration TestingResearchers Chunyi Zhang, Jin Zeng, and Xiaoqi Li released a comprehensive paper on system penetration testing that was first submitted on…

Global: Hardware Variations Enable Backdoors in Machine Learning ModelsA study released on arXiv on January 29, 2026 demonstrates that subtle numerical differences across computing hardware can be leveraged to embed…

Global: Study Reveals Black-Box Evasion Attacks Threaten In-Context Learning ClassifiersResearchers have introduced a new adversarial framework, ICL-Evader, that targets large language models (LLMs) used for in-context learning (ICL) text classification.…