Amazon: China-Linked Actors Repurpose Cairn AI Agents to Exploit MONAI Label Zero-Day
Amazon Threat Intelligence reports that China-linked actors have weaponized the open-source Cairn AI penetration testing framework to deploy autonomous, LLM-driven offensive agents. These agents operate without human intervention, performing reconnaissance, vulnerability discovery, and post-exploitation. In a significant escalation, the agents identified and exploited a zero-day vulnerability (referenced as CVE-2026-133968 in some sources) within MONAI Label, an open-source medical imaging AI platform. This exploitation resulted in root-level server takeover and unauthorized access to academic and medical research environments. The campaign signifies a critical shift from human-assisted AI tools to fully autonomous AI-driven cyberattacks capable of independent exploitation.