AI Ethics: A Guide to Ethical AI

Author: Alexandria Jacobson

Publisher: Builtin

Publication Year: 2022

Summary: The following article focuses on Apixio, an AI healthcare company that is focused on delivering actionable insights to improve health outcomes. Their AI ethics revolve around the idea that healthcare data collection must be held to a higher standard than other industry organizations. They have best practices outlined for AI ethics. First, one must know the impact of your product’s AI. A company must be able to articulate how they are planning to use AI and what impact it will have on its customers. Second, companies must establish company values and ensure that ethics guidelines align with those values. Once you can operationalize those ethics, a company will then be able to work that into its products. Third, a company must emphasize transparency. Sometimes AI algorithms become too complex, sometimes beyond the point of human understanding. Making sure there is transparency surrounding the algorithm can reduce potential biases and ensure that the customer base does not assume the worst when it comes to AI provided by the company. Fourth, companies must focus on eliminating human bias. One solution that this author provides to decrease biases is to have a checklist for engineers as they think through the data before building models. This checklist may include questions such as: How was the data collected? What is the history behind the data? Who collected it? Fifth, it is important for companies to have AI risk assessments and audits. From neutral third parties, these audits can point out potential risks in how a company’s AI is being used.