Watch the recording on YouTube
- The session opened with a brief introduction, highlighting the increasing significance of AI in insurance, its projected $1.1 trillion annual value, and its application areas, such as underwriting, fraud detection, and claims management.
- AI poses risks like reputational damage, legal liability, and challenges in a rapidly evolving regulatory environment.
- Armilla AI's role in the industry was introduced, emphasizing its tech-based AI assessments and risk transfer programs.
EU AI Act:
- Dennis Noordhoek, Director of Public Policy & Regulation at The Geneva Association detailed the EU AI Act's three-tier risk model (unacceptable, high, and low risk).
- Insurance falls under the high-risk category, especially in life and health sectors, requiring stringent regulatory compliance.
- The Act's timeline was discussed, with full implementation expected 24 months post-publication.
UK Regulatory Approach:
- Charlotte Clark, Director of Regulation at The Association of British Insurers described the UK's pro-innovation and sectoral approach.
- The UK's strategy focuses on integrating AI within existing regulatory frameworks rather than creating new cross-sectoral legislation.
- Emphasis on promoting innovation, consumer protection, and robust governance.
US Regulatory Landscape:
- Kathleen Birrane, Maryland Insurance Commissioner and NAIC Innovation, Cybersecurity and Technology Committee Chair and Lindsey Klarkowski, Director of Data Science & AI/ML Policy at The National Association of Mutual Insurance Companies explained the NAIC's model bulletin, promoting a principles-based approach to AI regulation.
- The US approach involves contextual education for regulators and establishing governance and risk management frameworks for AI use in insurance.
- Discussion on the challenges of dual regulation, the need for uniform standards, and ongoing efforts to address third-party data and model vendors.
Advice:
- Common themes included the importance of governance, risk management, and testing for bias and fairness.
- Significant differences remain in how jurisdictions handle AI regulation, impacting multinational insurers and insurtech vendors.
- Recommendations for companies included engaging in policy discussions, understanding jurisdictional nuances, and ensuring transparency and effective communication in AI practices.
Conclusion:
- The panel emphasized the need for ongoing collaboration and dialogue among regulators, insurers, and stakeholders to navigate the evolving AI regulatory landscape effectively.
- Building trust through transparency, robust governance, and continuous learning from various regulatory approaches were highlighted as critical for the industry's future.