6 papers
Security--Fidelity Tradeoffs: The Hidden Cost of Prompt Injection Defense
Mitchell Hermon, Rahul Gupta, Weitong Ruan +2
We identify a security-fidelity tradeoff in defending LLMs against indirect prompt injection: defenses resist injected instructions largely by suppressing untrusted text, which cor…
When Actions Go Off-Task: Detecting and Correcting Misaligned Actions in Computer-Use Agents
Yuting Ning, Jaylen Jones, Zhehao Zhang +5
Computer-use agents (CUAs) have made tremendous progress in the past year, yet they still frequently produce misaligned actions that deviate from the user's original intent. Such m…
SkillHarm: Lifecycle-Aware Skill-Based Attacks via Automated Construction
Yuting Ning, Zhehao Zhang, Yash Kumar Lal +8
Agent skills occupy a privileged position in the agent workflow, as agents are expected to implicitly follow and execute them, rendering third-party skills a vulnerable attack surf…
SWAN: Semantic Watermarking with Abstract Meaning Representation
Ziping Ye, Gourab Dey, Christos Christodoulopoulos +7
We introduce SWAN (Semantic Watermarking with Abstract Meaning Representation), a novel framework that embeds watermark signatures into the semantic structure of a sentence using A…
How Catastrophic is Your LLM? Certifying Risk in Conversation
Chengxiao Wang, Isha Chaudhary, Qian Hu +3
Large Language Models (LLMs) can produce catastrophic responses in conversational settings that pose serious risks to public safety and security. Existing evaluations often fail to…
Customize Multi-modal RAI Guardrails with Precedent-based predictions
Cheng-Fu Yang, Thanh Tran, Christos Christodoulopoulos +3
A multi-modal guardrail must effectively filter image content based on user-defined policies, identifying material that may be hateful, reinforce harmful stereotypes, contain expli…