#instructions


Malicious MCP Servers Can Split Instructions to Make AI Coding Agents Exfiltrate Secrets

Aug 11, 2026 // 17:31

A malicious tool server connected to an AI coding assistant can quietly walk off with SSH keys, environment secrets, source code, and customer data without […]

A clever font-rendering technique allows bad actors to conceal harmful instructions from artificial intelligence systems.

Mar 17, 2026 // 21:14

A new type of attack exploits how fonts are rendered to trick AI helpers into overlooking harmful commands embedded in web pages, using seemingly harmless […]