The system is designed to prevent model substitution, such as labs secretly swapping high-end models for cheaper ones, and to verify third-party claims about an AI's involvement in a task. However, the current release is an open-source research demonstration intended for further community development rather than a ready-to-deploy commercial product.