Third-party cyber evaluations involving OpenAI models

OpenAI explains recent third-party cybersecurity evaluation incidents and outlines new safeguards to strengthen AI model testing and evaluation.

Read on openai.com