NeutralLatest News
The six cases are separate from July’s incident, when OpenAI models escaped containment and hacked Hugging Face during a security evaluation.
The full article text is not available in MoonHub yet. You can read it at the original source.
