The shifting regulatory environment shapes the indexed insurance solutions market in the US.
The session highlighted the changing regulatory landscape and governance frameworks surrounding insurance solutions linked to proprietary and custom indices, with an emphasis on responsible AI governance.
Vidal Vanhoof, counsel derivatives & structured products at Davis Polk, led a training module during the SRP Americas 2026 Masterclass focused on the governance of indexed insurance solutions.
He started with the federal and state regulatory frameworks and shared that state insurance regulation is covered by the 1945 McCarran-Ferguson Act while the National Association of Insurance Commissioners (NAIC) is responsible for developing model rules and guidance.
“[The] key point here is that multiple regulatory regimes comply simultaneously to the same product, and among other things, that means that particular attention has to be paid to determine the correct framework for a particular product,” he said.
Participants of the SRP Americas 2026 Masterclass in Scottsdale, Arizona on 22 September.
Index oversight
Regulators remain focused on transparency of proprietary indices and custom indices that are used as underlyings, especially where limited historical data is available.
A key governance expectation is for market participants to uphold comprehensive index governance practices, ensuring transparent methodologies, well-defined discretion protocols, and appropriate controls for conflicts of interest, he noted.
“[Sales] literature used in connection with these products and related advertising must not contain materially receiving statements or omissions, including with respect to index methodology, discretion, and complexity risk,” said Vanhoof.
Regulators are concerned that newly created proprietary indices may be reverse-engineered to show favourable historical results.
Meanwhile, the NAIC is considering revisions to Model Regulation 245, covering illustration length, disclosures, accountability and non-guaranteed credited rates. Another rule, Model Regulation 275, aims to align insurance-agent standards with broker-dealer obligations, but state adoption varies.
“The idea behind Model Regulation 275 was to harmonise the requirements that apply at state level to insurance agents, to those that apply for broker dealers pursuant to the SEC's for API and impose a best interest standard in the following,” he said.
When applied to sophisticated or bespoke indices, the rules may result in heightened suitability obligations for market participants.
“[A] little bit of performance history, that tested results, caps participation rates, buffers, floors, and other more complicated features do need to be fully explained so that the customer can understand and assess the product,” Vanhoof explained.
Regulated AI
“[We]'re seeing that insurers are increasingly using AI in areas such as underwriting, pricing, customer service, claims handling, marketing, and fraud detection, whereas from the index creation perspective, index sponsors are also using AI to assist with index creation,” said Vanhoof.
Regulators have been ahead of the curve in shaping AI-related regulatory expectations for state adoption. The NACI issued a model bulletin focusing on the use of AI systems by insurers in December 2023.
According to Vanhoof, the regulation is designed as a reference framework to help align state insurance regulators' expectations for insurers using AI.
“It also goes on to remind insurance parties that decisions that ultimately impact consumers that are made or supported by advanced analytical and computational techniques, including AI, have to comply with all of typical insurance laws and regulations, including those related to unfair trade practices,” he said.
Vanhoof also pointed out that “ultimate legal and regulatory authority rests with the relevant institution and not the algorithm” and emphasised that product and index design should incorporate rigorous governance measures and meaningful human oversight to ensure appropriate management of AI-enabled processes.
Vanhoof concluded that the direction is clear with regulators expecting to see effective human involvement in how AI tools are deployed, particularly where they bring influences on customer outcomes.