Trojans in Artificial Intelligence (TrojAI) Final Report
• IARPA TrojAI program tackles AI trojan backdoor vulnerability in modern AI systems. • Found detection via weight analysis and trigger inversion techniques to identify hidden back
Ai Security
Infostealer Steals OpenClaw AI Agent Configuration Files and Gateway Tokens
• Infostealer variant of Vidar exfiltrated OpenClaw AI agent config files. • Stolen files include openclaw.json, device.json, soul.md with tokens, keys, operational principles. • T
The Next Frontier of Runtime Assembly Attacks: Leveraging LLMs to Generate Phishing JavaScript in Real Time
• Attackers embed a benign page that calls an LLM API to generate malicious JavaScript in real time. • Prompt engineering bypasses AI safety guardrails, producing polymorphic phish