SIR: Self-improving Red-teaming for Compute Use Agents

arXiv:2608.30207 · cs.CR, cs.AI · Submitted 2026-08-31 · Read on arXiv

cs.CR, cs.AI

Submitted: 2026-08-31

Updated: 2026-10-03

Terminology

• Analyzer — 21× in this paper · explained in 1 episode
• Benchmark — 10× in this paper · explained in 2 episodes
• LLM judge — 9× in this paper · explained in 2 episodes
• Attack Success Rate — 8× in this paper
• Benchmarking — 4× in this paper · explained in 3 episodes
• Robustness — 4× in this paper · explained in 5 episodes
• Ablation — 3× in this paper · explained in 1 episode
• Cross-Model Transfer — 3× in this paper · explained in 2 episodes
• Trace — 3× in this paper · explained in 2 episodes
• Verification — 3× in this paper · explained in 2 episodes
• AI Agents — 2× in this paper · explained in 2 episodes
• Attack Success Rate (ASR) — 2× in this paper · explained in 2 episodes
• LLM-as-a-Judge — 2× in this paper · explained in 3 episodes
• Adaptation — 1× in this paper · explained in 1 episode
• Adversarial Attacks — 1× in this paper · explained in 1 episode
• Adversarial Robustness — 1× in this paper · explained in 3 episodes
• Agent Hijacking — 1× in this paper · explained in 1 episode
• Agentic System — 1× in this paper · explained in 2 episodes
• Distillation — 1× in this paper · explained in 2 episodes
• Iterative Refinement — 1× in this paper · explained in 3 episodes
• Reasoning Traces — 1× in this paper · explained in 3 episodes
• Thirteenth International Conference on Learning Representations — 1× in this paper
• Vision-Language Models (VLMs) — 1× in this paper · explained in 6 episodes

Sources

• Jailbreaking Black Box Large Language Models in Twenty Queries
• How Vulnerable Are AI Agents to Indirect Prompt Injections? Insights from a Large-Scale Public Competition
• WASP: Benchmarking Web Agent Security Against Prompt Injection Attacks
• EIA: Environmental Injection Attack on Generalist Web Agents for Privacy Leakage
• RedTeamCUA: Realistic Adversarial Testing of Computer-Use Agents in Hybrid Web-OS Environments
• AutoDAN-Turbo: A Lifelong Agent for Strategy Self-Exploration to Jailbreak LLMs
• EVA: Evolving Semantic Adversaries for Red-Teaming GUI Agents Against Environmental Injection Attacks
• Tree of Attacks: Jailbreaking Black-Box LLMs Automatically
• Ignore Previous Prompt: Attack Techniques For Language Models
• UI-TARS: Pioneering Automated GUI Interaction with Native Agents
• Is LLM-as-a-Judge Robust? Investigating Universal Adversarial Attacks on Zero-shot LLM Assessment
• A Comprehensive Survey of Agents for Computer Use: Foundations, Challenges, and Future Directions
• MUZZLE: Adaptive Agentic Red-Teaming of Web Agents Against Indirect Prompt Injection Attacks
• On Adaptive Attacks to Adversarial Example Defenses
• OpenCUA: Open Foundations for Computer-Use Agents
• AgentVigil: Generic Black-Box Red-teaming for Indirect Prompt Injection against LLM Agents
• Dissecting Adversarial Robustness of Multimodal LM Agents
• CoP: Agentic Red-teaming for Large Language Models using Composition of Principles
• AdvAgent: Controllable Blackbox Red-teaming on Web Agents
• InjecAgent: Benchmarking Indirect Prompt Injections in Tool-Integrated Large Language Model Agents

Related papers

• SoK: AI-Augmented Binary Reversing
• Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
• Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
• Efficient Fuzzy PSI under One-Sided Assumptions
• Sealing the Audit-Runtime Gap for LLM Skills
• Token Composition: A Graph Based on EVM Logs