Third-party cyber evaluations involving OpenAI models (opens in a new tab)
OpenAI explains recent third-party cybersecurity evaluation incidents and outlines new safeguards to strengthen AI model testing and evaluation.
Read at OpenAI (opens in a new tab)


