Tackling Bias and Unfairness in AI