If not then it's second hand. Neither of the OpenAI or Anthropic announcements recently say much about their security and governance posture.