How MCP Servers Can Expose Enterprise Secrets
Summary: MCP servers, the middlemen that let AI agents reach enterprise tools and data, hold the credentials, service account keys and API tokens for everything they touch, and the article describes how those secrets leak through plaintext configuration files, over-permissioned access and prompt injection. It cites CVE-2025-6514 in mcp-remote, an OAuth proxy downloaded over 400,000 times, where a malicious server could trigger OS command injection and remote code execution on the machine running the proxy.
What you should know: A leaked secret in an MCP server does not just expose data; it lets an attacker act on the system, and because credentials are scattered across config files and environment variables with no central inventory, they rarely get rotated and stay valid indefinitely.
Read the original at The Hacker News ↗
- Topics
- AI Security
- Related
- SECURITY AFFAIRS AI-CYBERSECURITY NEWSLETTER ROUND 8 Cyber Risks · Nov 15, 2026Pentagon Renames CDAO ‘Super Intelligence Office’ Cyber Risks · Nov 15, 2026186 Fake AI Domains Redirect Users to Click Fix Lures, Browser Malware, and Search Hijackers Cyber Risks · Nov 12, 2026