Evasion Attacks

🤔 What Are Adversarial Attacks? Adversarial attacks exploit a fundamental characteristic of how AI models work: they learn decision boundaries