AI use cases
Find vulnerabilities in AI models and feed the findings back into training to improve model robustness.
From GPT-Red: Can AI red teams stop prompt injections? by IBM Technology