Skip to content
AImpact
IT EN
Safety Beginner Also known as: Test avversariale

Red teaming

A practice where a team actively tries to attack a model or an AI system, looking for jailbreaks, security holes, and dangerous uses, in order to find them before release.

ShareLinkedInX

In practice

AI labs do it in-house and with external experts before shipping a model. If you put AI in production do the same on your product: ask colleagues to break it before customers do. Even one rough hour beats the first public bug.

Related terms

Jailbreak Prompt injection Alignment ASL

Seen in the wild

4 entries mentioning it
  1. Promptfoo Red Teaming: open source automated red-teaming with CI integration and comparative benchmark
    Medium
  2. Garak: the open source vulnerability scanner for LLMs
    Medium
  3. MITRE ATLAS v2: the AI attack taxonomy updated with real case studies
    High
  4. Red Teaming LLMs with LLMs: the DeepMind paper that changed safety testing
    High
← All terms