There are 6 common attacks faced by modern AI systems: hinting attacks, training data extraction, backdoor manipulation of models, adversarial examples, manipulation of training data of models using data contamination attacks, and data leakage attacks.
