At OpenAI, we proactively adapt, including by building comprehensive security measures directly into our infrastructure and models.