Below is the thesis for a new, short paper I’m working on that proposes shifting away from a deontological approach to ethics, to more of a virtue-ethics approach, whereby ethics is treated as a set of skills instilled in people over time, as opposed to an immutable set of rules that must be followed.
We’ve all seen companies espouse their enduring principles of safety, security, transparency, and fairness. On the surface, these principles seem both reasonable and responsible, albeit somewhat anodyne. Who, after all, would argue against transparency? Many of the Responsible AI frameworks practiced today are based on a predefined set of ethical guidelines that serve as ideals employees are expected to maintain through an existing understanding of ethics and specialized analytic tools designed to identify and minimize unethical decision-making. A commonality among these frameworks is a deontological assumption that the employees will make ethical decisions based on the company’s predetermined rules and principles. What is often missing is an articulation of exactly how the employees will learn to make ethical decisions or how to improve their abilities over time. Therefore, instead of approaching ethics as a set of immutable principles that must be adhered to, we should instead treat ethics as a skill set cultivated and developed over time. This means shifting from a deontological approach to ethics to a virtue-ethics approach, opening the door to the notion of “ethics as a skillset” rather than a set of absolute rules or duties.