AI governance: Trust and transparency

AI is prevalent in every industry—everything from recommendation engines to tracking the spread of the COVID-19 virus. This application of AI empowers organizations to innovate and introduce efficiency, but it also introduces risks around how AI is employed to make business decisions.

In general, governance usually relates to who is responsible for data, how data is handled, who has access to see particular datasets in an enterprise, and how they get to use the data. This same notion of establishing a governance framework to protect the data and the consumers also applies to AI models, which may exhibit behavior that may be unfair or biased—or both—to some individuals or groups of people. Data governance has evolved over a period, and there are well-established methodologies and best practices, but we now need to do something similar for AI models. The challenge with traditional data governance is the increasing volume of...