New AI detects deepfakes by analyzing motion, not just faces

This new AI system from UC Riverside and Google can detect manipulations even when faces aren’t visible. By analyzing motion, background, and subtle inconsistencies, it tackles a growing challenge in misinformation and digital trust.

What implications do you see for content moderation, privacy, and fake video generation and detection?

Should this approach be used in social media platforms, or are there limitations we should consider?