Verification of possible exposure of sensitive data used in training machine learning models and LLMs. Includes prompt injection, membership inference attacks, and model inversion techniques to determine if confidential information, PII, or intellectual property present in training datasets can be extracted or deduced from model responses.