This session will cover the major concepts of the SOA report “Ethical Use of Artificial Intelligence for Actuaries,” through two examples, one from Life and one from P&C. Some primary sources of PA and AI risks include the following:
• Data risks, resulting from imperfect or misunderstood data
• Bias risks, resulting from both deliberate and unintentional bias that can manifest itself at many points in the PA or AI development process
• Automation risks, where the automation of a process can have unintended consequences
• Amateurish development, where the systems are not developed by AI professionals and unintended consequences result.
Central to the discussion is the concept of the Social Context and how it can guide when ethical questions arise. This session will also cover the actuarial precepts involved. Yes, you can get professionalism credit at this relevant and entertaining session.