Susceptibility of Continual Learning Against Adversarial Attacks