From the moment we wake until the time we turn off the phone at night, AI plays an important role in our lives: proposing healthy routines based on our sleeping habits, recommending news articles based on our past preferences, predicting traffic and suggesting routes to work, selecting the entertainment program for the evening. How do we make technology work for us and our society in a responsible way? AI ethics is no longer a topic of academic debates, it has become a business imperative. In this talk, I would like to discuss a range of biases contributing to fairness issues in the machine learning pipeline: historical, representation, measurement, human interaction, and evaluation biases. Using two-sided recommender systems as a specific context, a number of fairness principles will be presented along with some potential solutions to ensure fair and equitable outcomes for users of these AI systems.