Quantifying the Carbon Footprint of Adversarial Machine Learning: Balancing Robustness and Sustainability

The carbon emissions of adversarial machine learning models are directly correlated with their robustness against attacks, highlighting the need to integrate environmental considerations into secure ML system design.