#runtime-protection News & Analysis

2 articles tagged with #runtime-protection. AI-curated summaries with sentiment analysis and key takeaways from 50+ sources.

2 articles

AIBullisharXiv – CS AI · May 77/10

🧠

AgentTrust: Runtime Safety Evaluation and Interception for AI Agent Tool Use

AgentTrust is a runtime safety layer that intercepts AI agent tool calls before execution to prevent unsafe actions like accidental deletion, credential exposure, or data exfiltration. The system achieves 95-96.7% verdict accuracy across benchmarks using deobfuscation, risk chain detection, and LLM-based judgment, addressing a critical gap in AI agent safety infrastructure.

AINeutralarXiv – CS AI · Apr 147/10

🧠

ClawGuard: A Runtime Security Framework for Tool-Augmented LLM Agents Against Indirect Prompt Injection

Researchers introduce ClawGuard, a runtime security framework that protects tool-augmented LLM agents from indirect prompt injection attacks by enforcing user-confirmed rules at tool-call boundaries. The framework blocks malicious instructions embedded in tool responses without requiring model modifications, demonstrating robust protection across multiple state-of-the-art language models.