In addition to ensuring that ML models and applications are respectful and cognizant of people’s privacy, and that they don’t infringe on human rights by violating privacy laws, it’s also important that AI technologies be protected from cyberattacks. This involves building robust ML models, which is another fundamental part of responsible AI.
Building reliable and secure ML systems is important to the success of any ML model/system, and is known as robustness. Robustness measures the stability of an algorithms’ performance when a model ...
