Comment by jnovacho

Sure, but are they running audio recognition and image recognition models on each upload just to classify it?