Both Anthropic and OpenAI had to delay their rollouts in order to add more safeguards.

With these safeguards in place, supposedly the incident we are discussing would not have taken place.