Tag
3 articles
OpenAI's internal AI red-teaming model, GPT-Red, outperformed human red-teamers 84% to 13% on prompt injection tests and discovered novel attack techniques.
NVIDIA's garak framework offers a comprehensive solution for defensive LLM red-teaming, enabling organizations to identify vulnerabilities and enhance model safety.
OpenAI has acquired Promptfoo, an open-source AI red-teaming tool used by over 125,000 developers and 30 Fortune 500 companies. The move strengthens OpenAI's focus on AI application security and will integrate into its new Frontier enterprise platform.