Exposure of Critical Vulnerabilities in OpenClaw: 360's Vulnerability Mining AI Agent Reveals New Security Challenges for AI-Native Applications
360's AI agent discovered and reported 3 vulnerabilities in OpenClaw (1 high-risk, 2 medium-risk), all fixed. This marks a shift from rule-based to intelligent AI security auditing, enhancing AI application safety. The high-risk flaw involved local script approval and execution mechanisms.....