Tech companies are advancing in artificial intelligence (AI) at a rapid pace, and the insurance industry is keenly adopting these innovations. AI applications in insurance range from customer service, underwriting, claims processing, to fraud detection. This AI infusion promises insurers both enhanced customer experience and profitability through faster response times and better decision accuracy.
However, the insurance sector is rigorously regulated. So, regulators are striving to understand AI’s incorporation by insurers to craft appropriate monitoring guidelines. On July 17, the National Association of Insurance Commissioners (NAIC) unveiled a draft titled “Model Bulletin on the Use of Algorithms, Predictive Models, and Artificial Intelligence Systems by Insurers.”
This draft isn’t a regulation, but a communication template for regulators. Its goal? To guide insurers in AI use while adhering to market conduct, corporate governance, and fair trade practices. It also aims to strike a balance between nurturing innovation and protecting consumers from potential AI-related harm, such as discrimination or bias.
The Emergence of the NAIC H Committee
The NAIC established the Innovation Cybersecurity and Technology (H) Committee, previously known as Innovation and Technology (EX) Task Force, to navigate tech innovations in insurance. In 2019, this committee birthed the Big Data and AI Working Group to evaluate AI’s rise, its implications on consumer protection, market dynamics, and regulatory framework. By 2021, this group began assessing how insurers use AI and the control measures in place.
Key figures leading the H Committee include Maryland Insurance Commissioner Kathleen Birrane, Colorado’s Michael Conway, and Iowa’s Doug Ommen. These leaders, backed by a consortium of state representatives, spearheaded the draft bulletin’s development.
To gauge AI’s application in insurance, the H Committee performed three AI usage surveys covering auto insurance, homeowners, and life insurance.
Key Highlights of the Draft Bulletin
The draft delineates what regulators expect from insurers using AI. It advocates for a documented program based on risk assessment of AI deployment, ensuring adherence to laws, including trade practices.
Main areas covered include:
- AI system program guidelines.
- Risk management and controls.
- Third-party AI system considerations.
The draft clarifies terms such as “AI systems” and provides documentation guidelines covering governance, risk management, and usage. It also elaborates on the documentation type insurers should maintain, from policies to training resources.
According to this bulletin, AI systems encompass predictive models and algorithms. It’s advised that insurers document their oversight methods and compliance controls. Additionally, insurers are accountable for third-party AI solutions they employ and must ensure these align with regulations.
The draft positions insurance regulators to inspect insurers’ AI applications. Insurers should anticipate inquiries on their governance, risk management, and third-party AI engagements. The bulletin offers a flexible approach, allowing insurers various methods to demonstrate compliance.
Potential industry feedback might revolve around:
- Definitions clarity.
- Incorporation of existing laws.
- Document confidentiality.
- Expertise requirements of insurance examiners.
- Documentation timeline.
Public Engagement on the Draft
Feedback on the draft is open until Sept. 5. The H Committee will also entertain in-person comments during the NAIC summer meeting on Aug. 13. Those wishing to participate should inform the NAIC by Aug. 9.