In the wake of recent high-profile AI bias scandals, companies have begun to realize that they need to rethink their AI strategy to include not just AI Fairness, but also Algorithmic Fairness more broadly as a fundamental tenet. At the Pragmatic Institute, we educate Fortune 500 companies about data science and AI. Through our work, we’ve discovered that many companies struggle to form a clear definition of algorithmic fairness for their organization. Without a clear definition, well-meaning fairness initiatives languish in the realm of good intentions and never arrive at meaningful impact. But defining fairness is not as easy as it may seem. Two examples highlight just how challenging this can be.