See how a regional mutual insurer operationalized AI governance
For Insurance

Audit-ready AI governance for insurers

Swept helps insurers put AI under real governance: supervise every model in production and turn oversight into evidence your examiners can review. Then expand into cost control, reliability, private model access, and custom AI solutions, all on the same governed foundation.

Book a demoExplore the governance hub

AI introduces both opportunity and risk for insurers

Regulatory and financial exposure

Incorrect decisions in underwriting, pricing, or claims can trigger regulatory scrutiny, legal exposure, or financial loss.

Inconsistent or opaque AI behavior

Insurers need predictable, explainable AI decisions across diverse customer cohorts, not demo-level accuracy.

Complex real-world inputs

Dialect, incomplete information, ambiguous events, historical notes, scanned documents, and long conversational threads all increase variance in AI behavior.

The regulatory reality

Regulators have already responded

Across the country, insurance regulators have moved on AI. Most states now expect insurers to govern AI under the laws already on the books, and to show their work when examined.

Existing laws already apply

Unfair trade practice, unfair discrimination, and rating laws apply to any decision an AI system touches. Regulators are not waiting for new legislation.

A documented program is expected

Insurers are expected to maintain a documented AI governance program covering oversight, risk management, third-party models, and the documentation that proves it.

Examinations want evidence

Regulators increasingly ask for proof that controls were implemented, tested, and enforced: model inventories, validation records, change approvals, and a data-to-decision trail.

State Adoption Tracker

Where the NAIC model bulletin has been adopted

25 jurisdictions have adopted the model bulletin and 4 more have insurance-specific AI guidance. Hover a state for details. Status as of April 1, 2026; confirm with each state's department of insurance.

Adopted (25)
Insurance-specific guidance (4)
No state action
Alaska: AdoptedAlabama: No state actionArkansas: AdoptedArizona: No state actionCalifornia: Insurance-specific guidanceColorado: Insurance-specific guidanceConnecticut: AdoptedDelaware: AdoptedFlorida: No state actionGeorgia: No state actionHawaii: AdoptedIowa: AdoptedIdaho: No state actionIllinois: AdoptedIndiana: No state actionKansas: No state actionKentucky: AdoptedLouisiana: No state actionMassachusetts: AdoptedMaryland: AdoptedMaine: No state actionMichigan: AdoptedMinnesota: No state actionMissouri: No state actionMississippi: No state actionMontana: No state actionNorth Carolina: AdoptedNorth Dakota: No state actionNebraska: AdoptedNew Hampshire: AdoptedNew Jersey: AdoptedNew Mexico: No state actionNevada: AdoptedNew York: Insurance-specific guidanceOhio: No state actionOklahoma: AdoptedOregon: No state actionPennsylvania: AdoptedRhode Island: AdoptedSouth Carolina: No state actionSouth Dakota: No state actionTennessee: No state actionTexas: Insurance-specific guidanceUtah: No state actionVirginia: AdoptedVermont: AdoptedWashington: AdoptedWisconsin: AdoptedWest Virginia: AdoptedWyoming: No state actionDistrict of Columbia: Adopted

Adopted the model bulletin

  • Alaska, Bulletin B 24-01, Feb 1, 2024
  • Arkansas, Bulletin 13-2024, Jul 31, 2024
  • Connecticut, Bulletin No. MC-25, Feb 26, 2024
  • Delaware, Domestic and Foreign Bulletin No. 148, Feb 5, 2025
  • District of Columbia, Bulletin 24-IB-002-05/21, May 21, 2024
  • Hawaii, Insurance Commissioner Memorandum No. 2025-13A, Dec 10, 2025
  • Illinois, Company Bulletin 2024-08, Mar 13, 2024
  • Iowa, Insurance Division Bulletin 24-04, Nov 7, 2024
  • Kentucky, Bulletin No. 2024-02, Apr 16, 2024
  • Maryland, Bulletin No. 24-11, Apr 22, 2024
  • Massachusetts, Bulletin No. 2024-10, Dec 9, 2024
  • Michigan, Bulletin 2024-20-INS, Aug 7, 2024
  • Nebraska, Insurance Guidance Document No. IGD-H1, Jun 11, 2024
  • Nevada, Bulletin 24-001, Feb 23, 2024
  • New Hampshire, Bulletin Docket #INS 24-011-AB, Feb 20, 2024
  • New Jersey, Insurance Bulletin No. 25-03, Feb 11, 2025
  • North Carolina, Bulletin No. 24-B-19, Dec 18, 2024
  • Oklahoma, Bulletin No. 2024-11, Nov 14, 2024
  • Pennsylvania, Insurance Notice 2024-04, Apr 6, 2024
  • Rhode Island, Insurance Bulletin No. 2024-03, Mar 15, 2024
  • Vermont, Insurance Bulletin No. 229, Mar 12, 2024
  • Virginia, Administrative Letter 2024-01, Jul 22, 2024
  • Washington, Technical Assistance Advisory 2024-02, Apr 22, 2024
  • West Virginia, Insurance Bulletin No. 24-06, Aug 9, 2024
  • Wisconsin, Insurance Bulletin, Mar 18, 2025

Insurance-specific AI guidance

  • California, Bulletin 2022-5
  • Colorado, 3 CCR 702-10
  • New York, Insurance Circular Letter No. 7
  • Texas, Commissioner's Bulletin B-0003-26
Start with governance

Make your AI audit-ready

Swept builds the governance infrastructure and turns it into a board-ready Trust Report. It is where most insurers begin, and it makes everything else they build defensible.

What the governance foundation delivers

  • A board-ready Trust Report that maps every AI system to a defensible governance program
  • Every AI interaction logged, every change tracked, every decision reviewable
  • Risk appetite, thresholds, and role-based approvals enforced in production
  • Audit trails and explainability bundles your examiners can open on request
Explore GovernanceState-by-state hub
01

Monitor

Swept captures inputs, outputs, reasoning, and tool calls across the AI behind underwriting, quoting, policy servicing, fraud checks, and claims, and keeps a live inventory of the models in use.

02

Evaluate

Every recommendation is tested against underwriting guidelines, claims rules, fraud signals, regulatory policy, and the risk appetite you define.

03

Control

When a response is out of policy, Swept blocks it, routes it to a human reviewer, or falls back to a safer path, and records every step as evidence.

Customer proof
“A governance framework and a governance program are not the same thing.”

Regional mutual insurance carrier · AI governance engagement

5

Claims agents shipped

5

States monitored daily

L1

Agent autonomy, by design

Read the case study
The platform

Five offerings, built to support insurers

Governance is the foundation. The same governed, LLM-agnostic environment powers the rest, so what your team runs and what Swept builds for you is supervised the same way.

Cost Optimization

Right-size AI across your workforce. Swept meters spend across every model and every employee, flags where a cheaper or purpose-built model would do the job better, and keeps usage inside budget.

Explore Cost Optimization

Governance

Audit-ready oversight and a board-ready Trust Report: every AI interaction logged, every change tracked, every decision reviewable.

Explore Governance

Reliability

Accuracy thresholds, hallucination limits, and drift alarms on the models behind risk scoring, claims, and fraud, so failures are caught before they reach a customer or a payout.

Explore Reliability

Private AI Access

Use any frontier model while policyholder PII stays inside your boundary. Redaction, masking, and VPC or on-prem deployment match your security posture.

Explore Private AI Access

Workflow Solutions

Swept engineers build insurer-specific AI on your real claims, underwriting, and servicing workflows, integrated with your core systems and audit-ready from day one.

Explore Workflow Solutions
Use cases

Where insurers put Swept to work

Claims automation

Identify unsafe or incorrect recommendations, misrouted claims, bad summarization logic, and drift in classification patterns.

Fraud detection workflows

Monitor for false positives, shifting thresholds, and degradation in anomaly-detection behavior across customer cohorts.

Underwriting assistants

Check for divergence from underwriting guidelines, inconsistent risk scoring, or policy-breaking suggestions.

Customer service and FNOL agents

Ensure policy explanations, eligibility checks, and intake guidance stay accurate, compliant, and consistent.

Broker, agent, and adjuster copilots

Guarantee that internal AI tools provide reliable reasoning, follow documentation standards, and respect privacy rules.

Built to hold up under examination

Most states now expect insurers to govern AI, with 25 adopting formal expectations and 4 more issuing AI guidance, as of April 1, 2026. Start with governance, then expand on a foundation examiners can review.

Book a demoEstimate your AI risk
Enterprise security

Enterprise security designed for insurers

Full data-privacy guardrails

Optional redaction, PII masking, and no logging of sensitive inputs where required.

Audit trails and reporting

Evidence built for compliance, legal, and risk review.

Deployment flexibility

Cloud, VPC, or on-prem to match your enterprise security posture.

Learn the basics

Insurance AI resources

Plain-language guides to AI governance, financial examinations, and what regulators actually expect.

Guide

What is the NAIC Model Bulletin on AI?

The NAIC Model Bulletin on the Use of AI Systems by Insurers is the template most states use to set AI governance expectations. Here is what it says and why it matters.

Article

Insurance AI Governance Demands More Than a Checklist

Insurance carriers approve AI at one speed and govern it at another. Until governance becomes infrastructure, that gap will keep producing the failures policies were designed to prevent.

Article

AI in Financial Examinations: What Regulators Will Ask and What Carriers Must Produce

State insurance examiners are adding AI-specific inquiries to financial and market conduct examinations. Continuous supervision generates examination-ready evidence as a byproduct of normal operations.

Article

AI Is a Catalyst for Insurance. Governance Needs to Keep Pace.

The insurance industry is adopting AI as a catalyst for transformation. But catalysts without governance create uncontrolled reactions. Insurance needs AI governance that leads adoption, not governance that chases it.

Guide

AI in Insurance: Key Regulatory Definitions

The NAIC Model Bulletin defines the terms that carry legal weight, from AI System to Adverse Consumer Outcome to Model Drift. Here is what each one means for insurers.

Guide

What is an AIS Program?

An AI Systems Program (AIS Program) is the written program the NAIC Model Bulletin expects every insurer to maintain. Here are its four pillars and what each one requires.

Frequently asked questions

How does Swept help with AI compliance and examinations?
Swept builds the governance infrastructure regulators expect and turns it into a board-ready Trust Report: model inventories, validation records, change approvals, and a clear data-to-decision trail. It supervises your models in production and keeps that evidence examination-ready. Our state-by-state hub tracks where AI governance requirements have landed.
What does Swept AI do for insurers beyond compliance?
Governance is the foundation. On the same governed environment, Swept controls AI cost across your workforce, sets reliability thresholds on the models you run, keeps policyholder data private across any model, and builds custom AI solutions on your real claims and underwriting workflows.
Can Swept run inside our security boundary?
Yes. Swept deploys in the cloud, in a VPC, or on-prem, with redaction and PII masking so policyholder data stays inside your environment while you use the models you choose.
Which AI systems can Swept govern?
Claims automation, fraud detection, underwriting assistants, FNOL and customer service agents, and internal broker, agent, and adjuster copilots, across whichever models and vendors power them.
How do we get started?
Most insurers start with governance. Book a demo to see Swept produce governance evidence against your own rules, or run the AI risk assessment to estimate your exposure first.

Start with governance. Grow from there.

Make your AI audit-ready, supervise it in production, and expand across the platform on one governed foundation.

Book a demoRun the risk assessment