AI regulation in insurance varies by state. Colorado leads with AI governance laws, while other states are developing frameworks.
TL;DR
AI insurance regulation varies by state: Colorado has the most comprehensive AI governance laws (bias testing, documentation, audits), Connecticut requires transparency, New York and California have issued guidance. The NAIC Model Bulletin provides a framework adopted by 3 states with 15+ considering adoption. Multi-state agents should comply with the strictest standard (Colorado) for simplicity and future-proofing. Expect 10-15 more states to regulate AI in insurance by 2028.
What Is State AI Insurance Regulation?
State AI insurance regulation encompasses state-specific laws governing the use of artificial intelligence in insurance operations — including bias testing (Colorado), transparency requirements (Connecticut), consumer disclosure (California), and impact assessments (Virginia) — with the NAIC Model Bulletin providing a framework for nationwide adoption.
Key Data Points
- Colorado leads US AI insurance regulation with mandatory bias testing, documentation, and audit requirements — 15+ states are considering similar frameworks based on the NAIC Model Bulletin
- Multi-state insurance agents should comply with the strictest AI regulation (Colorado's standard) for simplicity and future-proofing as 10-15 additional states are expected to adopt AI rules by 2028
AI regulation in insurance is fragmented. Some states have enacted specific AI legislation. Others are developing guidance. Most are watching and waiting. Here is the state-by-state landscape as of 2026.
States with Active AI Insurance Regulation
Colorado
Status: Most comprehensive AI insurance regulation in the US
Colorado SB 21-169 and subsequent rules require:
- Testing AI systems for unfair discrimination before deployment
- Documenting AI decision-making processes
- Regular audits of AI outputs for bias
- Governance frameworks for all AI-assisted insurance decisions
- Penalties for AI systems that produce discriminatory outcomes
What agents need to do:
- Verify that AI tools used in Colorado have been tested for bias
- Document which AI tools are used and for what purposes
- Maintain records of AI-assisted recommendations
- Ensure AI quoting tools do not produce discriminatory pricing outcomes
Connecticut
Status: AI transparency requirements enacted
Connecticut requires:
- Disclosure when AI is used in insurance decisions affecting consumers
- Consumer right to request human review of AI-assisted decisions
- Documentation of AI systems used in underwriting and pricing
What agents need to do:
- Inform Connecticut clients when AI is used in their insurance process
- Offer human review options for AI-assisted recommendations
- Maintain records of AI disclosure
New York
Status: AI guidance issued, regulation pending
New York DFS has issued guidance on AI in insurance:
- AI systems must comply with existing unfair discrimination laws
- Insurers must be able to explain AI-driven decisions to regulators
- AI models used in underwriting must be validated for accuracy and fairness
What agents need to do:
- Ensure AI tools can explain their recommendations if questioned
- Verify that AI quoting and recommendation tools have been validated
- Follow existing unfair discrimination requirements
California
Status: Consumer protection rules for AI in insurance
California's approach focuses on consumer protection:
- AI-driven decisions must not violate the Unfair Insurance Practices Act
- Consumers have the right to know if AI influenced their insurance pricing
- AI systems processing personal information must comply with CCPA/CPRA
What agents need to do:
- Comply with CCPA/CPRA requirements for AI data processing
- Be prepared to disclose AI usage to clients upon request
- Ensure AI tools do not produce unfairly discriminatory outcomes
Virginia
Status: AI impact assessment requirements
Virginia requires:
- Impact assessments for high-risk AI systems in insurance
- Reasonable measures to prevent algorithmic discrimination
- Consumer notification when AI significantly impacts insurance decisions
States Developing AI Frameworks
Active Development
- Texas — Considering AI governance requirements for insurance
- Florida — Reviewing AI impact on property insurance pricing
- Illinois — Developing AI transparency requirements
- Ohio — Studying AI bias in insurance underwriting
- Georgia — Considering AI disclosure requirements
Monitoring / No Action
Most remaining states are monitoring federal developments and the NAIC Model Bulletin before enacting state-specific regulations.
The NAIC Model Bulletin
The National Association of Insurance Commissioners' Model Bulletin on AI provides a framework that states can adopt:
Key Provisions
- Governance — Organizations using AI must have documented oversight policies
- Risk management — AI systems must be assessed for adverse outcomes
- Transparency — Consumers must be informed when AI affects them
- Fairness — AI must be tested for unfair discrimination
- Human oversight — AI decisions must have human review mechanisms
- Third-party AI — Organizations are responsible for AI tools they use, even if built by vendors
Adoption Status
As of 2026:
- Adopted the Model Bulletin framework: Colorado, Connecticut, Virginia
- Partially adopted: New York, California
- Under consideration: 15+ states
- Not yet addressed: Remaining states
Compliance Strategies for Multi-State Agents
Strategy 1: Comply with the Strictest Standard
If you sell in Colorado, Connecticut, AND states without AI regulation, apply Colorado's standards everywhere. This is:
- Simplest to manage (one set of practices)
- Future-proof (as other states adopt similar rules)
- Defensible (exceeds minimum requirements in every state)
Strategy 2: State-Specific Configuration
For large agencies operating in many states, configure AI tools by state:
- Colorado leads: Full bias testing, documentation, and disclosure
- Connecticut leads: AI transparency and human review options
- Other states: Follow NAIC Model Bulletin as best practice
Strategy 3: Product-Specific Approach
- Medicare: Follow CMS rules nationally (supersede state AI rules)
- Health/ACA: Follow state-specific rules plus federal ACA requirements
- Life/Annuity: Follow state-specific suitability rules plus AI governance
- P&C: Follow state-specific rating and underwriting AI rules
What to Expect in 2027-2028
Federal AI Regulation
Congress is considering federal AI legislation that may:
- Preempt some state regulations
- Establish national standards for AI in financial services (including insurance)
- Create federal AI audit requirements
State Expansion
Expect 10-15 additional states to adopt AI insurance regulations by 2028, likely based on the NAIC Model Bulletin framework.
Enforcement Ramp-Up
States with existing AI regulations will move from guidance to enforcement:
- Colorado will begin active audits of AI systems in insurance
- Fines for non-compliance will set precedents
- First AI-related insurance enforcement actions will establish case law
Action Items for Insurance Agents
- Know your states — Identify which states you sell in and their current AI requirements
- Document your AI tools — List every AI tool you use, what it does, and how you verify its outputs
- Configure compliance guardrails — AI instructions must include state-specific compliance boundaries
- Maintain audit trails — All AI interactions should be logged and accessible for regulatory review
- Stay current — AI regulation is evolving rapidly; review state requirements quarterly
- Choose compliant vendors — Verify that AI tool vendors have tested for bias, maintain transparency, and support audit requirements
The regulatory landscape is complex and changing. But the core principle is simple: use AI responsibly, document everything, and maintain human oversight of all AI-assisted decisions.
From Experience
- The agents who will be best positioned when their state enacts AI regulation are the ones who started documenting their AI usage and verification processes before it was required. Start now — retroactive compliance is much harder than proactive compliance.
Frequently Asked Questions
Which states regulate AI in insurance?
As of 2026, Colorado has the most comprehensive AI insurance regulation (bias testing, documentation, audits). Connecticut requires AI transparency and human review. New York and California have issued guidance. Virginia requires impact assessments. 15+ states are considering adoption of the NAIC Model Bulletin framework.