Skip to content
← Glossary

AI & ML

AI red-teaming

Adversarial testing specifically aimed at surfacing a model’s unsafe, biased, or exploitable behaviors — distinct from general security red-teaming.