A new paper from researchers in Italy and Germany proposes a method to detect deepfake videos based on biometric face and voice behavior, rather than artifacts created by face synthesis systems, expensive watermarking solutions, or other more unwieldy approaches. The framework requires an input o