Monitor-Validity Failures: LLM Safety Evidence Paths at Evolving Framework Boundaries
Modern LLM applications rely on runtime monitors that inspect tool calls, traces, streams, detector outputs, and verifie
#cs.AI
共收录 1652 条相关情报。
← 返回所有主题MobileInject: Evaluating Reasoning-to-Action Security of MLLM-Based Mobile Agents under Cross-Modal Consistency Injection
Multimodal large language model (MLLM)-based mobile agents are rapidly evolving from conversational assistants into auto
StratFL: Countering Covert Model Poisoning Attacks with Strategic Federated Learning
Advanced persistent threats (APTs) are a major concern in federated learning (FL) deployments in critical sectors such a
Private Retrieval Augmented Generation via Random Projection
Retrieval-Augmented Generation (RAG) enhances the capabilities of large language models (LLMs) by querying external stru
Language Is a Security Surface for Agent Guards
Step-level guards decide, before each tool call, whether an agent's next action is safe given the user's request and the
Shortcut Sensitivity in Agent-Safety Evaluations
AI agents take actions through tools. Some actions are sensitive, so guards that supervise them must themselves be evalu
Encoded but Exposed: Language Embeddings can Reveal More Than They Should in Recommender Systems
Language embeddings are increasingly used in recommenders to represent item content, but their privacy risks remain poor
BiShield-TEE: On the (In-)Security of Unilateral Weight Obfuscation in On-Device TEE-Shielded LLM Partition
On-device deployment of large language models (LLMs) is increasingly adopted due to latency and privacy requirements, bu
Matoi: Selective Revalidation of Consequence Evidence for Agent Authorization
Matoi separates reusable consequence evidence from current execution authority and durable occurrence history. Its finit
FedSumUp: User-Auditable Privacy-Aware Representation Sharing for Resource-Constrained Federated Intelligence
Existing federated data condensation methods face privacy risks, as uploaded synthetic data may preserve sensitive infor
CARE-SD: Explainable Reasoning Architecture for Local SLM Self-Disclosure Analysis in Textual Data
Cybersecurity interviews uncover critical security experiences, behaviors, and privacy concerns often not fully captured
Stronger Judges Do Not See More: Deployment-Integrity Blindness in LLM Evaluation
LLM-agent evaluation reports behavioral scores but leaves the identity of the deployed runtime stack unreported. A tampe
Approve What, Exactly? Effect Binding and Trusted Rendering for Agent Step-Up Authorization
AI agents now pause for approval before they send money, delete files or email outsiders, and that approval is usually t
Screwtape: Adversarial Transcript Editing for Monitor Red-Teaming
We present Screwtape, a pipeline that converts real agent transcripts into attacks that evade action-only monitors. An L
INFUSE: A Unified Surface Benchmark for Indirect Prompt Injection in Agentic Systems
LLM-based agents increasingly execute long-horizon, end-to-end workflows involving shell commands, code modification, de
From Training to Escalation: Forensic Analysis and Local Recreation of RL/RLHF-Induced Privilege Escalation in AI Agents (CVE-2026-65616)
Mid-2026 reporting on autonomous AI evaluation incidents claimed that frontier large language model (LLM) agents escaped
Counterfactual Neighborhood Risk Fields: Evidence-Grounded Selective Safety for Vision-Language Models
Representation-based safety defenses for vision-language models (VLMs) often learn to separate harmful from benign input
P2Patch: Exploit-Guided Vulnerability Repair
Large language models are increasingly deployed as autonomous agents for automated program repair, patching software vul
GeoCloak: An Agentic Framework for Semantics-Preserving Geolocation Privacy via Peripheral Clue Manipulation
The rapid advancement of large vision language models (LVLMs) has made geolocation inference from images highly accurate
CoRedSkill: Stealthy Skill Composition Attacks via AST-Guided Injection
Agents extend their capabilities through reusable skills that package instructions, executable code, and resources. Comp