AI, Ethics, and Insurance: Why Bias Begins at the Data Source

Video

AI governance expert Sean Tumanov explores the ethical, legal, and practical risks of generative AI in insurance -from data privacy to model bias - and the critical importance of transparency, collaboration, and early testing in responsible model development.

Apr 09, 2025

Shawn Tumanov, Director, Data, Model and AI Governance Executive, GEICO

Tags: AI and Technology (including Fintech) Insurance Model risk

The views and opinions expressed in this content are those of the thought leader as an individual and are not attributed to CeFPro or any other organization

Sean Tumanov, an AI governance leader with deep experience across banking and insurance, outlines the complex ethical and operational risks of deploying AI in financial services.

He emphasizes that responsible AI begins with understanding the origin and purpose of data, ensuring transparency and explainability, and proactively testing for bias - especially when models impact individuals.

With rising regulatory expectations and consumer trust at stake, he highlights the need for early-stage governance, clear use-case justification, and industry-wide collaboration to align on best practices and protect stakeholders.

Shawn Tumanov Bio

Shawn Tumanov is the Model & AI Governance Executive at GEICO. He is responsible for establishing an enterprise framework for identifying, measuring and mitigating Model & AI Risks. Prior to this role, Shawn was the Director of Enterprise Data & Analytics at BMO, where he implemented an efficient process for advancing the AI/ML practices, focused on streamlined governance program. With an extensive background spanning over two decades in financial industry, Shawn began his career as bank examiner at the Office of the Comptroller of the Currency (OCC) and has since progressed to hold various senior roles in risk management.