Detecting Adversarial Texts In Ai Systems