Anthropic CEO Uncertain How Claude Aided School Strike
Anthropic CEO Dario Amodei admitted he does not know exactly how Claude AI was used in a missile strike that killed around 120 children at an Iranian elementary...
AI SECURITY THREAT
3 articles
Malicious Code Hides in Test Files, Evades Scanners
Anthropic Skill scanners pass every security check while missing malicious code hidden in test files. Researcher Jeevan Jutla from Gecko Security revealed that ...
OpenClaw Flaw Gave Attackers Silent Admin Access
A critical vulnerability in OpenClaw, the viral AI agentic tool with 347,000 GitHub stars, allowed attackers with the lowest-level permissions to silently escal...