Until now, efforts to identify and combat deepfakes and AI scams have been scattered. Which proposed standard will dominate?