In March 2026, cybersecurity researchers at Zscaler ThreatLabz uncovered a sophisticated malware campaign targeting the rapidly growing ecosystem of agentic AI workflows. Threat actors weaponized OpenClaw, a viral open-source framework previously known as Clawdbot and Moltbot, to deploy remote access trojans and information stealers. By publishing a deceptive plugin named “DeepSeek-Claw,” attackers successfully tricked autonomous […]
The post Malicious OpenClaw Skill Targets DeepSeek Agentic AI Workflows appeared first on Cyber Security News.