As an advocate for ethical AI, I strongly believe that frontier-model developers should publish their model cards before commercial release. Transparency is crucial in ensuring accountability and understanding the limitations and biases of AI models. This helps mitigate harm and builds public trust in technology.