In this webinar, Bob Bryan and John Valutkevich will discuss the concept of AI explainability (also referred to as interpretability) and how it fosters trust among users—particularly in healthcare.

The discussion clarifies the key concept of source attributes, making it more accessible to the audience by connecting it to familiar ideas (such as model cards). It also addresses how industry standards and regulatory requirements for AI explainability are coalescing key concepts that enable AI trust for end users.

Key Themes: 

• Explainable AI – AI users, especially in healthcare, need to understand how AI makes decisions before they can trust and adopt it. 
• Source Attributes & Model Cards – These concepts help describe AI models, much like a “nutrition label” for AI, outlining intended uses, risks, and limitations. 
• Challenges in AI Adoption – Clinicians struggle to trust AI because they may not know the breadth of the data models used in its development and the impact on risk management 
• Industry Trends & Regulations – The webinar will tie predictive decision support interventions (pDSI) into emerging regulations and broader industry efforts to improve AI transparency and trust.