“Pindrop's AI technology for voice identification analyzes 8,000 samples per second to prevent deepfakes.”