Md Jahangir Alam 0003, Tanzim Ahad, Ismail Hossain, Sai Puppala, Yoonpyo Lee, Syed Bahauddin Alam, Sajedul Talukder. HALLPERM: Exposing the Safety Illusion in LLM Tool Use via Implicit Privilege Escalation and Semantic Risk. In Christopher W. Geib, Ron P. A. Petrick, editors, Proceedings of the 2026 AAAI Summer Symposium Series, Seoul, South Korea, June 22-24, 2026. pages 238-245, AAAI Press, 2026. [doi]
Abstract is missing.