Researchers expose Claude Cowork sandbox escape, Anthropic disagrees
Security firm Armadin detailed an attack chain achieving root-level code execution inside Claude Cowork's sandbox on Windows, bypassing network restrictions and...
GENERATIVE AI SECURITY VULNERABILITY
4 articles
200,000 MCP Servers Vulnerable to Command Execution Flaw
OX Security researchers discovered a critical architectural flaw in Anthropic's Model Context Protocol, the AI agent communication standard adopted by OpenAI an...
Microsoft Copilot Studio Patch Can't Stop Data Leaks
Microsoft patched CVE-2026-21520, a CVSS 7.5 indirect prompt injection vulnerability in Copilot Studio, discovered by Capsule Security and deployed January 15. ...
GrafanaGhost flaw silently stole enterprise data via AI
A vulnerability dubbed GrafanaGhost allowed attackers to silently exfiltrate sensitive enterprise data through Grafana's AI features without phishing, user appr...