generative ai systems across modalities, ranging from text, image, audio, and video, have broad social impacts, but there exists no official standard for means of evaluating those impacts and which impacts should be evaluated. We move toward a standard approach in evaluating a generati