MetaBackdoor: Exploiting Positional Encoding as a Backdoor Attack Surface in LLMs
Authors: Rui Wen, Mark Russinovich, Andrew Paverd, Jun Sakuma, Ahmed Salem | Published: 2026-05-14
PickleFuzzer: A Case Study in Fuzzing for Discrepancies Between Python Pickle Implementations
Authors: Justin Applegate, Andreas Kellas | Published: 2026-05-14
Toward Securing AI Agents Like Operating Systems
Authors: Lukas Pirch, Micha Horlboge, Patrick Großmann, Syeda Mahnur Asif, Klim Kireev, Thorsten Holz, Konrad Rieck | Published: 2026-05-14
EVA: Editing for Versatile Alignment against Jailbreaks
Authors: Yi Wang, Hongye Qiu, Yue Xu, Sibei Yang, Zhan Qin, Minlie Huang, Wenjie Wang | Published: 2026-05-14
Defenses at Odds: Measuring and Explaining Defense Conflicts in Large Language Models
Authors: Xiangtao Meng, Wenyu Chen, Chuanchao Zang, Xinyu Gao, Jianing Wang, Li Wang, Zheng Li, Shanqing Guo | Published: 2026-05-14
Exploiting LLM Agent Supply Chains via Payload-less Skills
Authors: Xinyu Liu, Yukai Zhao, Xing Hu, Xin Xia | Published: 2026-05-14
Watermarking Game-Playing Agents in Perfect-Information Extensive-Form Games
Authors: Juho Kim, Fei Fang, Tuomas Sandholm | Published: 2026-05-14
Identifying AI Web Scrapers Using Canary Tokens
Authors: Steven Seiden, Triss Ren, Caroline Zhang, Taein Kim, Enze Liu, Emily Wenger | Published: 2026-05-13
AIエージェントによる悪用に関する脅威
本記事では、OWASP Foundationによる「OWASP Top 10 for Agentic Applications 2026 」に記載されている脅威やその対策について最新の研究動向を交えながら解説します。特に、本記事では AIエージェントによる悪用に関する2つの脅威を扱います。
Model-Agnostic Lifelong LLM Safety via Externalized Attack-Defense Co-Evolution
Authors: Xiaozhe Zhang, Chaozhuo Li, Hui Liu, Shaocheng Yan, Bingyu Yan, Qiwei Ye, Haoliang Li | Published: 2026-05-13