Instances of bias and discrimination across many machine learning units have lifted many moral questions regarding the usage of artificial intelligence. How can we safeguard versus bias and discrimination if the training data alone may very well be produced by biased human processes? Even though companies usually have good intentions for their auto