One argument against publishing adversarial robustness research is that it might make your systems easier to attack.
One argument against publishing adversarial robustness research is that it might make your systems easier to attack.