ABSTRACT
The transformative and transitional nature of technology creates an implicit sense of risk which generally requires a reasonable degree of caution—its newness raises questions regarding how to appropriately monitor and control it to protect against the unknown. Though ongoing advances may seem unprecedented, history has given us numerous examples of unanticipated ethical risks encountered with technology that empowers data collection, analysis, and manipulation. The relatively young field of data science has a story filled with explosive growth and vital problems solved, but also with ethical pitfalls. A cursory glance reveals successive examples of how data and tools assumed to be safe and innocuous revealed themselves to carry unintended risks, invisible until after the fact. A brief consideration of a few examples reveals how the ethical risks encountered in data science and AI are often easy to miss, and subject to rules of discourse and how they enable emergent risks to be addressed when they exceed the imagination. Solving these problems carries challenges for both individual data scientists and the organizations on whose behalf they act, requiring both individual ethical agency and sustainable, responsible practices.
