June 9, 2026
The Wharton School
Philadelphia, PA

This conference brings together leaders from academia, industry, and regulation to explore how AI is reshaping the future of insurance. Taking place on June 9, 2026, at the Wharton School in Philadelphia, the workshop will examine critical issues at the intersection of AI, analytics, and insurance—from fairness and transparency in pricing models to the growing challenges of climate risk and disaster insurance.

The program features a keynote panel on fairness, transparency, and trust in insurance pricing, with perspectives from leading experts and industry practitioners. Throughout the day, academic and industry sessions will highlight advances in responsible machine learning, regulatory expectations, and real-world applications of AI in insurance.

Event Webpage

REGISTER HERE

— Agenda at a Glance —

Subject to change


8:30 a.m.–9:00 a.m.
Registration and Welcome Coffee

9:00 a.m.–9:10 a.m.
Opening Remarks

9:10 a.m.–10:40 a.m.
Keynote Panel: When AI Meets Insurance: Regulation, Accountability, Climate, and Affordability

Home insurance premiums are surging, insurers are retreating from high-risk markets, and AI is reshaping how risk gets priced, often in ways that are hard to see and harder to challenge. What does responsible AI look like in insurance? Who holds it accountable? And as climate losses mount, who can still afford to be covered and how to solve the affordability crisis? Our opening keynote brings together a regulator, a climate finance researcher, an actuarial scientist, and an AI ethics expert for a frank conversation about the forces pulling insurance markets apart and what it will take to hold them together.

10:40 a.m.–11:10 a.m.
Morning Tea Break

11:10 a.m.–12:10 p.m.
Session I: Bias, Uncertainty, and Causality in Machine Learning

Machine learning models can be powerful, but can we trust what they appear to tell us? Getting interpretation right starts with precision: being clear about what you are actually looking for before choosing a method. This session examines popular interpretation strategies, including partial dependence plots and distillation trees, and the pitfalls that come with them. It then tackles two deeper layers of uncertainty: uncertainty about what an interpretation reveals given the model, and uncertainty about the model itself.

Moving from interpretation to causation introduces a further layer of complexity. Even when a policy change shows a net benefit on average, it may still harm a sizable subpopulation. Because we never observe counterfactuals, the exact extent of that harm is fundamentally unknowable. This session also explores how tight bounds on subpopulation-level harm can be derived and estimated robustly, and what this means for evaluating the fairness of algorithmic decisions in practice.

12:10 p.m.–1:10 p.m.
Lunch and Networking

1:10 p.m.–2:10 p.m.
Session II: Fairness for Insurance Pricing

What do we really mean by fairness and bias in insurance pricing? How should insurers, regulators, and actuaries evaluate fairness in practice — and what are the implications for consumers and firms?

This session brings together industry and academic perspectives to examine the evolving landscape of fair insurance pricing across both general and life insurance. The discussion will explore the challenges of balancing predictive accuracy, regulatory compliance, transparency, and equity in modern pricing models.

2:10 p.m.–2:40 p.m.
Break

2:40 p.m.–3:40 p.m.
Session III: AI and Insurance Market Failures

Insurance markets are limited by several demand- and supply-side frictions, including information asymmetries, model risk and uncertainty, market power, and regulatory frictions that may inhibit product innovation. As AI becomes more central to insurer business strategies, an important question emerges: to what extent can these technologies ease or exacerbate these existing market frictions?

This session will discuss the role of AI in insurance pricing, risk modeling, monitoring, and information, while also examining how regulation may need to evolve as AI-driven tools become increasingly embedded across the industry.

3:40 p.m.–4:10 p.m.
Break

4:10 p.m.–5:10 p.m.
Workshop Session IV: Regulating AI in Insurance Markets

Insurance markets are comprehensively governed by a complex web of state laws and regulations. As insurers increasingly incorporate AI into nearly every facet of their operations, from underwriting and claims handling to marketing and sales, they are raising novel and difficult questions for state insurance regulators. These questions include how traditional prohibitions on unfair discrimination should apply to algorithmic decision-making, what forms of explainability should be required in underwriting decisions, and how regulators should respond to the risk that AI-driven marketing may manipulate consumers.

This session will provide an overview of how state insurance law and regulation are beginning to adapt to these challenges, how they are likely to evolve in the coming years, and how economic and empirical research can and should help shape that trajectory.

It will also explore the practical and analytical tools available for evaluating AI systems in insurance and informing regulatory policy. Particular attention will be given to fairness metrics, statistical and machine-learning approaches to fair decision-making, the welfare consequences and trade-offs associated with fair and accountable regulation for insurers and consumers, and testing and auditing frameworks designed to assess fairness, detect potential biases, and support regulatory compliance and oversight.

5:10 p.m.–5:20 p.m.
Closing Remarks