Chernoff information is compared to Kullback-Leibler divergence for privacy metrics in adversarial classification.