Microsoft and OpenAI claim the model is highly secure, but that they’re always improving and adding security features and ...