An attack that corrupts a model's training data so the resulting model behaves the way the attacker intended.
Continue to AI University →