Cover Image

When Convenience Becomes Risk: A Semantic View of Under-Specification in Host-Acting Agents

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Lu, Di, Liao, Yongzhi, Mu, Xutong, Zheng, Lele, Cheng, Ke, Dong, Xuewen, Shen, Yulong, Ma, Jianfeng
Format:	Preprint
Published:	2026
Subjects:	Cryptography and Security Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Constraining Host-Level Abuse in Self-Hosted Computer-Use Agents via TEE-Backed Isolation
by: Lu, Di, et al.
Published: (2026)

PriFFT: Privacy-preserving Federated Fine-tuning of Large Language Models via Hybrid Secret Sharing
by: You, Zhichao, et al.
Published: (2025)

EBCC: Enclave-Backed Confidential Containers via OCI-Compatible Runtime Integration
by: Lu, Di, et al.
Published: (2026)

Arca: A Lightweight Confidential Container Architecture for Cloud-Native Environments
by: Lu, Di, et al.
Published: (2026)

Differentially Private Subspace Fine-Tuning for Large Language Models
by: Zheng, Lele, et al.
Published: (2026)

The Trust Paradox in LLM-Based Multi-Agent Systems: When Collaboration Becomes a Security Vulnerability
by: Xu, Zijie, et al.
Published: (2025)

Local Differential Privacy is Not Enough: A Sample Reconstruction Attack against Federated Learning with Local Differential Privacy
by: You, Zhichao, et al.
Published: (2025)

When Safety Becomes a Vulnerability: Exploiting LLM Alignment Homogeneity for Transferable Blocking in RAG
by: Li, Junchen, et al.
Published: (2026)

Enhancing Privacy of Spatiotemporal Federated Learning against Gradient Inversion Attacks
by: Zheng, Lele, et al.
Published: (2024)

When Understanding Becomes a Risk: Authenticity and Safety Risks in the Emerging Image Generation Paradigm
by: Leng, Ye, et al.
Published: (2026)

Adaptive Backdoor Attacks with Reasonable Constraints on Graph Neural Networks
by: Dong, Xuewen, et al.
Published: (2025)

When Everyday Devices Become Weapons: A Closer Look at the Pager and Walkie-talkie Attacks
by: Sarker, Pantha Protim, et al.
Published: (2025)

When AIOps Become "AI Oops": Subverting LLM-driven IT Operations via Telemetry Manipulation
by: Pasquini, Dario, et al.
Published: (2025)

AsyncSC: An Asynchronous Sidechain for Multi-Domain Data Exchange in Internet of Things
by: Yang, Lingxiao, et al.
Published: (2024)

When Developer Aid Becomes Security Debt: A Systematic Analysis of Insecure Behaviors in LLM Coding Agents
by: Kozak, Matous, et al.
Published: (2025)

Guard-GBDT: Efficient Privacy-Preserving Approximated GBDT Training on Vertical Dataset
by: Song, Anxiao, et al.
Published: (2025)

Extracting Spatiotemporal Data from Gradients with Large Language Models
by: Zheng, Lele, et al.
Published: (2024)

No Attack Required: Semantic Fuzzing for Specification Violations in Agent Skills
by: Li, Ying, et al.
Published: (2026)

Patch2QL: Discover Cognate Defects in Open Source Software Supply Chain With Auto-generated Static Analysis Rules
by: Wang, Fuwei, et al.
Published: (2024)

When Denoising Becomes Unsigning: Theoretical and Empirical Analysis of Watermark Fragility Under Diffusion-Based Image Editing
by: Gu, Fai, et al.
Published: (2026)

Systematic Categorization, Construction and Evaluation of New Attacks against Multi-modal Mobile GUI Agents
by: Yang, Yulong, et al.
Published: (2024)

When Backdoors Go Beyond Triggers: Semantic Drift in Diffusion Models Under Encoder Attacks
by: Chen, Shenyang, et al.
Published: (2026)

Bridging Cloud Convenience and Protocol Transparency: A Hybrid Architecture for Ethereum Node Operations on Amazon Managed Blockchain
by: Hossain, S M Mostaq, et al.
Published: (2025)

Evolving Deception: When Agents Evolve, Deception Wins
by: Ying, Zonghao, et al.
Published: (2026)

When Skills Lie: Hidden-Comment Injection in LLM Agents
by: Wang, Qianli, et al.
Published: (2026)

Exposing LLM User Privacy via Traffic Fingerprint Analysis: A Study of Privacy Risks in LLM Agent Interactions
by: Zhang, Yixiang, et al.
Published: (2025)

When Ads Become Profiles: Uncovering the Invisible Risk of Web Advertising at Scale with LLMs
by: Chen, Baiyu, et al.
Published: (2025)

IstGPT: LLM-based Anomaly Detection for Spatial-Temporal Graph in Industrial Systems
by: Zhang, Yuchen, et al.
Published: (2026)

Under the Hood of SKILL.md: Semantic Supply-chain Attacks on AI Agent Skill Registry
by: Saha, Shoumik, et al.
Published: (2026)

A Large-Scale Analysis of Phishing Websites Hosted on Free Web Hosting Domains
by: Roy, Sayak Saha, et al.
Published: (2022)

AgentVisor: Defending LLM Agents Against Prompt Injection via Semantic Virtualization
by: Ying, Zonghao, et al.
Published: (2026)

Demystifying Cookie Sharing Risks in WebView-based Mobile App-in-app Ecosystems
by: Zhang, Miao, et al.
Published: (2025)

Threat-Specific Risk Assessment for IP Multimedia Subsystem Networks Based on Hierarchical Models
by: Shaikh, Abdullah Ehsan, et al.
Published: (2025)

Shuffling for Semantic Secrecy
by: Chen, Fupei, et al.
Published: (2025)

ACRFence: Preventing Semantic Rollback Attacks in Agent Checkpoint-Restore
by: Zheng, Yusheng, et al.
Published: (2026)

Red-Teaming LLM Multi-Agent Systems via Communication Attacks
by: He, Pengfei, et al.
Published: (2025)

Is It Really You Who Forgot the Password? When Account Recovery Meets Risk-Based Authentication
by: Büttner, Andre, et al.
Published: (2024)

TRUST: A Toolkit for TEE-Assisted Secure Outsourced Computation over Integers
by: Zhao, Bowen, et al.
Published: (2024)

SpatialJB: How Text Distribution Art Becomes the "Jailbreak Key" for LLM Guardrails
by: Mou, Zhiyi, et al.
Published: (2026)

WIPI: A New Web Threat for LLM-Driven Web Agents
by: Wu, Fangzhou, et al.
Published: (2024)