Viral AI Assistant OpenClaw Exposes Major Security Flaws as Cisco Finds 26% of AI Agent Skills Contain Vulnerabilities
Summary
Viral AI assistant OpenClaw exposes critical security flaws by executing dangerous commands without protection, while Cisco discovers 26% of AI agent skills contain vulnerabilities that enable data theft and bypass safety measures undetected.
Key Points
- OpenClaw, a viral open-source personal AI assistant, poses major security risks by running shell commands, executing scripts, and storing persistent memory while lacking built-in security protections
- Cisco researchers discover that 26% of 31,000 analyzed AI agent skills contain vulnerabilities, with malicious skills capable of data exfiltration and bypassing safety guidelines without user awareness
- Cisco releases an open-source Skill Scanner tool to help detect malicious AI agent skills that could introduce shadow AI risks and bypass traditional enterprise security controls