WEBINAR | Highmark Responsible AI Framework

Michael Barber, Director, Data Science R&D, Highmark Health. This webinar discusses considerations for a data-centric approach to machine learning (ML) and artificial intelligence (AI) development. It emphasizes the importance of human oversight, ethical data acquisition, and addressing potential biases in data. Michael also explores various scenarios that ML models and AI may encounter, including feature importance and counterfactual reasoning. Additionally, he highlights the significance of transparency throughout the analytics lifecycle and introduces the Responsible AI (RAI) framework at Highmark Health. The framework encompasses diverse recruitment practices, ethical decision-making, and a culture that promotes RAI principles. Michael also emphasizes ongoing research, regular discussions to recognize bias in datasets, and collaboration with Enterprise Risk and Governance to expand the Enterprise Risk Framework to explicitly address AI risks. The webinar concludes by discussing the establishment of an Artificial Intelligence Governance Committee and the implementation of various governance mechanisms, including AI Vendor Questionnaires, Use Case Risk Assessments, and Model Cards. Finally, Michael introduces a risk scoring system to quantify and prioritize risks associated with AI use cases.