Basic safety and Robustness: AI devices is usually prone to adversarial assaults, the place modest, imperceptible adjustments on the input may cause the program to produce erroneous or maybe risky choices. Making certain the protection and robustness of AI techniques is really a significant problem.