Threat entity extracted from intelligence sources
The entity you're looking for doesn't exist.
Multi-agent AI systems are vulnerable to privilege abuse when agents delegate tasks across complex chains. This risk, classified by OWASP as Identity & Privilege Abuse, can lead to agents exceeding their original…