No Attacker Needed: Unintentional Cross-User Contamination in Shared-State LLM Agents
No Attacker Needed: Unintentional Cross-User Contamination in Shared-State LLM Agents explores This research identifies and quantifies a critical security flaw in shared-state LLM agents where benign user interactions can unintentionally corrupt outcomes for other users, requiring new artifact-level defenses.. Commercial viability score: 4/10 in LLM Agents.
Updated 6 hours ago
BUILDER'S SANDBOX
Build This Paper
Use an AI coding agent to implement this research.
OpenAI CodexAI Agent
Lightweight coding agent in your terminal.
Claude CodeAI Agent
Agentic coding tool for terminal workflows.
AntiGravity IDEScaffolding
AI agent mindset installer and workflow scaffolder.
CursorIDE
AI-first code editor built on VS Code.
VS CodeIDE
Free, open-source editor by Microsoft.
Recommended Stack
Startup Essentials
Estimated $10K - $14K over 6-10 weeks.
See exactly what it costs to build this -- with 3 comparable funded startups.
7-day free trial. Cancel anytime.
Discover the researchers behind this paper and find similar experts.
7-day free trial. Cancel anytime.
Related Papers
- Builds On ThisZombie Agents: Persistent Control of Self-Evolving LLM Agents via Self-Reinforcing InjectionsViability: 1.0
- Builds On ThisFrom Storage to Steering: Memory Control Flow Attacks on LLM AgentsViability: 3.0
- Builds On ThisAgentDrift: Unsafe Recommendation Drift Under Tool Corruption Hidden by Ranking Metrics in LLM AgentsViability: 2.0
- Builds On ThisDon't Trust Stubborn Neighbors: A Security Framework for Agentic NetworksViability: 2.0
- Prior WorkSilent Egress: When Implicit Prompt Injection Makes LLM Agents Leak Without a Trace