PI-Hunter: Automated Red-Teaming for Exposing and Localizing Prompt Injections
Authors: Pengfei He, Lesly Miculicich, Vishesh Sharma, Ash Fox, George Lee, Jiliang Tang, Tomas Pfister, Long T. Le | Published: 2026-06-10
OCELOT: Inference-Leakage Budgets for Privacy-Preserving LLM Agents
Authors: Jin Xie, Songze Li | Published: 2026-06-10
Mind your key: An Empirical Study of LLM API Credential Leakage in iOS Apps
Authors: Pinran Gao, Lingxiang Wang, Ying Zhang, Fan Yang | Published: 2026-06-10
Categorical Robustness Assessment for Machine Learning based Network Intrusion Detection Systems
Authors: Mayank Raj, Nathaniel D. Bastian, Lance Fiondella, Gokhan Kul | Published: 2026-06-10
AIエージェント特有の脅威
AI技術の発展に伴い、人間に変わって特定のタスクを自律的に行うAIエージェントを用いたシステム（AIエージェントシステム）の利活用が期待されています。大規模言語モデル（Large Language Model、 LLM）を中核に、Chain-of-Thought、メモリ（短期・長期記憶）、LLM単体では実現できない機能を提供するツールといった要素技術を組み合わせたAIエージェントシステムは、その自律性と能力の高さから、様々なタスクの遂行が可能です。一方で、AIエージェントシステムは、従来型のアプリケーションとは異なる、
Online Shift Detection and Conformal Adaptation for Deployed Safety Classifiers
Authors: Jun Wen Leong | Published: 2026-06-10
Grammar-Constrained Decoding Can Jailbreak LLMs into Generating Malicious Code
Authors: Yitong Zhang, Shiteng Lu, Jia Li | Published: 2026-06-10
Can Open-Source LLM Agents Replace Static Application Security Testing Tools? An Empirical Assessment
Authors: Derek Yohn, Luke Flancher, Mirajul Islam, Khaled Slhoub | Published: 2026-06-10
Dummy Backdoor as a Defense: Removing Unknown Backdoors via Shared Internal Mechanisms for Generative LLMs
Authors: Kazuki Iwahana, Masaru Matsubayashi, Takuma Koyama, Toshiki Shibahara, Kenichiro Omintato, Akira Ito | Published: 2026-06-10
Defense Against Prompt Inversion Attacks: An Information-Theoretic Approach for LLM Collaborative Inference
Authors: Sayedeh Leila Noorbakhsh, Hossein Khalili, Nader Sehatbakhsh | Published: 2026-06-10