AI SECURITY • 354 / 397
Understand how untrusted text, tools and autonomy create new attack paths.
Malicious Tool
A malicious tool is an integration intentionally designed to steal data, mislead the model or perform harmful actions.
Think of it like
Think of Malicious Tool as an input-trust or privilege problem: words can influence software decisions, so boundaries must be explicit.
Real life
Email, documents and web pages can become hostile inputs to AI systems through attacks involving Malicious Tool.
SRE lens
Only connect trusted, reviewed MCP servers to privileged agents.
Remember thisA malicious tool is an integration intentionally designed to steal data, mislead the model or perform harmful actions.