社会ガイド

保険における AI に関する NAIC モデル速報

4 分で読めます
最終更新日 2026年9月15日

このページでは4 分で読めます

概要
ディープダイブ
戦略的影響
The Future of The NAIC Model Bulletin on AI in Insurance
現実世界の実装
リスクとガードレール
実装ロードマップ
探検を続けましょう
よくある質問

概要

It does not create a new law; it explains how existing insurance laws against unfair discrimination and unfair practices apply when decisions are made or supported by AI. It matters because state regulators can ask for that program during examinations, so it shapes how carriers, vendors and the agents who use carrier tools handle AI.

ディープダイブ

The bulletin, formally titled the Model Bulletin on the Use of Artificial Intelligence Systems by Insurers, was adopted by the NAIC in December 2023. The NAIC is an association of state regulators, so its model bulletin has no force until an individual state insurance department issues it, often with light edits. Within about two years, roughly half of US jurisdictions had adopted it or a close version, including Alaska, Connecticut, Illinois, New Hampshire, Pennsylvania and Vermont. Some states took different routes: Colorado has its own statute and regulations on external consumer data and algorithms, and New York's Department of Financial Services issued a separate circular letter in 2024 on AI and external data in underwriting and pricing. The core expectation is an AIS Program: a written program for the responsible use of AI systems, designed to reduce the risk of Adverse Consumer Outcomes, meaning decisions that adversely affect consumers in ways that violate insurance law. The program should cover governance (clear accountability, often up to senior management or the board), risk management and internal controls across the AI lifecycle, data practices, validation and testing, and oversight of third-party models and data. It applies to generative AI as well as predictive models, and it is meant to be proportionate: controls should scale with how much a system affects consumers. A common misconception is that the bulletin bans particular AI uses or sets numerical fairness thresholds. It does neither. It relies on existing statutes such as the Unfair Trade Practices Act and the Unfair Claims Settlement Practices Act, and it describes what documentation regulators may request. Another misconception is that buying a vendor tool shifts responsibility to the vendor. The bulletin is explicit that insurers remain responsible for outcomes from third-party systems they use. For agents, the bulletin is addressed to insurers, but it affects the tools carriers provide to producers and the questions agents should expect about how those tools are used.

技術的な洞察

In practice, an AIS Program usually starts with an inventory: every model or AI-enabled tool, its owner, purpose, data inputs, and where it touches consumers such as marketing, underwriting, pricing, claims or fraud. Each entry gets a risk tier based on consumer impact and degree of automation, which determines required controls. High-tier systems typically need documented model development, validation by someone independent of the builder, testing for unfair discrimination or proxy effects, drift monitoring, and a human review or override path. Generative AI adds controls for accuracy of outputs, data leakage and record retention of prompts and responses. For third parties, insurers add due diligence questionnaires, contractual audit and cooperation rights, and ongoing performance monitoring. The bulletin also lists categories of information regulators may request in an investigation, such as policies, inventories, validation results and vendor contracts, so teams design documentation to be producible on request rather than reconstructed later.

戦略的影響

リスクと安全性

AI による壊滅的な被害も日常的な被害も、誰がリスクを理解し、誰が行動できるかにかかっています。

より明確な判決

国民と専門家のリテラシーは、強力な安全政策が政治的に可能かどうかを左右します。

誇大広告を打ち破る

明確な説明は、誇大広告、研究室の PR、曖昧な倫理劇場に囚われることを減らします。

The Future of The NAIC Model Bulletin on AI in Insurance

More states are likely to consider adopting the bulletin, and regulators have discussed whether some expectations should eventually move into a model law or regulation with firmer requirements, particularly for third-party data and model vendors. Examination practice will probably matter as much as the text: as departments begin asking for AIS Program documents, insurers will learn what evidence examiners consider adequate. Federal policy on AI remains unsettled, and state approaches still differ, so carriers operating nationally will need to reconcile the NAIC framework with state-specific rules such as those in Colorado and New York.

現実世界の実装

A property insurer that uses a machine learning model to flag claims for fraud review documents who approved the model, what data it uses, and how often it is tested for drift, so it can hand that record to examiners during a market conduct exam.

A life insurer licensing a third-party underwriting score adds contract clauses giving it audit rights and requiring the vendor to cooperate with regulator inquiries, as the bulletin's third-party section recommends.

A mid-sized auto carrier classifies its customer-service chatbot as lower risk than its pricing model, applying lighter controls to the chatbot and formal bias testing to the model, reflecting the bulletin's proportionate approach.

An independent agency using a carrier's AI quoting portal learns that the carrier, not the agency, is responsible for the model's governance, but still trains staff to escalate outputs that look wrong or unfair.

リスクとガードレール

能力が複雑になる一方で、実存的なリスクを SF として扱います。
高度な自律性の下での調整による表面製品の安全性を混乱させる。
英語以外や専門家ではない聴衆には、低品質の情報源しか提供されません。

実装ロードマップ

製品の危害、誤使用、制御不能/調整不良のリスクを分離します。
どのような証拠がタイムラインと重大度についてのあなたの見方を変えるかを尋ねてください。
マーケティング上の主張よりも、一次情報源と具体的な評価を優先します。
意識だけでなく、キャリア、政策、資金、スキルなど、行動経路を 1 つ特定します。

探検を続けましょう

Free newsletter

Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.

One email each weekday. Unsubscribe in one click. We never sell or share your address.

Test yourself

Take the The NAIC Model Bulletin on AI in Insurance quiz

Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.

クイズを開始する

Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation

よくある質問

What is The NAIC Model Bulletin on AI in Insurance?

The NAIC Model Bulletin on the Use of Artificial Intelligence Systems by Insurers is guidance, adopted by the National Association of Insurance Commissioners in December 2023, that tells insurers to maintain a written, risk-based program governing how they build, buy and use AI. It does not create a new law; it explains how existing insurance laws against unfair discrimination and unfair practices apply when decisions are made or supported by AI. It matters because state regulators can ask for that program during examinations, so it shapes how carriers, vendors and the agents who use carrier tools handle AI.

How does the NAIC Model Bulletin on AI gain legal effect in a given state?

The NAIC is an association of state regulators. Its model bulletin only applies in a state once that state's insurance department issues it.

What does the bulletin call the written program insurers are expected to maintain for AI use?

The bulletin expects a written AIS Program covering governance, risk management, controls and third-party oversight for AI systems.

Under the bulletin, what is an Adverse Consumer Outcome?

Adverse Consumer Outcomes are decisions that adversely affect consumers in ways that are inconsistent with insurance laws, such as unfair discrimination.

Which statement about the bulletin's legal basis is accurate?

The bulletin does not create a new statute or thresholds; it relies on existing insurance laws and describes expectations under them.

A carrier buys an AI underwriting score from a vendor. According to the bulletin, who remains responsible for consumer outcomes from that score?

The bulletin is explicit that insurers remain responsible for outcomes from third-party systems they use, even when a vendor supplied the model.

学び続ける