State regulators are watching AI in insurance closely. Here is how to adopt AI responsibly — with guardrails, audit trails.
TL;DR
The NAIC and states like Colorado are implementing AI governance rules for insurance. Agents must ensure AI tools provide audit trails, human oversight, fairness testing, and transparency. unLocked CRM addresses these requirements with full communication logging, DNC checks, Review mode, custom AI instructions, and SOC 2-aligned infrastructure with Type II certification in progress.
What Is AI Governance in Insurance?
AI governance in insurance refers to the frameworks, policies, and oversight mechanisms that ensure artificial intelligence tools used by insurance agents and carriers operate transparently, fairly, and in compliance with state and federal regulations — including the NAIC Model Bulletin and Colorado SB 21-169.
Key Data Points
- The NAIC Model Bulletin requires AI governance frameworks, risk management, transparency, fairness testing, and human oversight for AI in insurance
- Colorado SB 21-169 is the first state law requiring AI fairness testing and decision documentation specifically for insurance
- Insurance agents using AI tools need audit trails, human-in-the-loop review mechanisms, and documented governance policies
AI adoption in insurance is accelerating. So is regulatory attention. The NAIC (National Association of Insurance Commissioners) has issued AI governance guidelines, Colorado has passed AI-specific insurance legislation, and multiple states are developing frameworks for AI oversight.
Insurance agents who adopt AI tools need to understand the regulatory landscape — not to avoid AI, but to use it responsibly.
The Regulatory Landscape (2026)
NAIC Model Bulletin on AI
The NAIC's Model Bulletin on the Use of Artificial Intelligence Systems by Insurers requires:
- Governance frameworks — Organizations using AI must have documented policies for AI oversight
- Risk management — AI systems must be assessed for potential adverse outcomes
- Transparency — Consumers must be informed when AI is used in decisions affecting them
- Fairness testing — AI outputs must be tested for unfair discrimination
- Human oversight — AI-assisted decisions must have human review mechanisms
Colorado SB 21-169 (AI Insurance Governance)
Colorado was the first state to legislate AI governance specifically for insurance:
- Requires insurers and intermediaries to test AI systems for unfair discrimination
- Mandates documentation of AI decision-making processes
- Requires regular audits of AI outputs for bias
- Imposes penalties for AI systems that produce discriminatory outcomes
State-Level AI Regulations
Multiple states are developing AI-specific insurance regulations:
- New York: AI guidance for underwriting and pricing
- California: Consumer protection rules for AI-driven insurance decisions
- Connecticut: AI transparency requirements for insurance communications
- Virginia: AI impact assessment requirements
What This Means for Insurance Agents
1. AI-Assisted Quoting
AI quoting tools that retrieve carrier rates and compare products are generally low-risk from a regulatory perspective because they present actual carrier data — not AI-generated recommendations.
unLocked's AI Quoting Suite retrieves real carrier rates from 1,252 integrations. The AI organizes and presents the data; the agent makes the recommendation. This human-in-the-loop model aligns with regulatory expectations.
2. AI-Generated Communications
AI Chatbot and AI-generated emails require more careful governance:
- Disclosure: Some states may require disclosure that a message was AI-generated
- Accuracy: AI-generated insurance information must be accurate — inaccurate information could constitute misrepresentation
- Record keeping: All AI-generated communications should be logged with full audit trails
unLocked CRM logs every AI-generated message with the context used to generate it, the DNC check result, and delivery confirmation — creating a comprehensive compliance record.
3. AI Voice Calling
Agent AI and Arwyn (inbound AI) handle actual conversations with prospects and clients. Compliance considerations include:
- Call recording disclosures in two-party consent states
- TCPA compliance for outbound AI calls
- Insurance solicitation rules — AI cannot make binding coverage recommendations without agent involvement
4. AI Lead Scoring
Predictive lead scoring raises fairness concerns if scoring models inadvertently discriminate based on protected characteristics (race, gender, age, disability). AI lead scoring models should:
- Not use protected characteristics as input variables
- Be tested for disparate impact across demographic groups
- Be documented and auditable
The Agent's AI Compliance Checklist
Before Adopting an AI Tool
- Does the tool log AI decisions? — Without audit trails, you cannot demonstrate compliance
- Is there human oversight? — Can you review and override AI outputs?
- What data does the AI use? — Understand inputs to assess fairness risk
- Where is data stored? — Ensure compliance with state data protection requirements
- Does the vendor have an AI governance policy? — Ask for documentation
Ongoing Compliance
- Review AI-generated communications monthly — spot-check for accuracy and tone
- Monitor lead scoring outcomes — check for demographic skew in scoring results
- Update AI instructions when regulations change
- Document your AI governance process — create a written policy even if not yet required
- Train staff on AI oversight — everyone who uses AI tools should understand their compliance responsibilities
The Competitive Advantage of Responsible AI
Agents who adopt AI responsibly — with documented governance, audit trails, and human oversight — will have a competitive advantage as regulations tighten:
- Trust: Clients increasingly want to know how their data is being used
- Compliance readiness: When your state adopts AI regulations, you are already compliant
- E&O protection: Documented AI governance strengthens your position in any dispute
- Carrier relationships: Carriers prefer agents who use technology responsibly
unLocked CRM's Built-In Compliance Features
- Full audit trails on all AI-generated communications
- DNC checks before every automated message
- Human-in-the-loop modes (Review mode for SMS, approval workflows for AI calls)
- Custom AI instructions that enforce compliance boundaries
- Daily reply caps to prevent over-communication
- Call recording compliance with two-party consent handling
- Data encryption at rest and in transit
- SOC 2 aligned infrastructure (Type II certification in progress)
From Experience
- The agents who proactively document their AI governance practices — even before regulations require it — are building a compliance foundation that competitors will scramble to create after the fact.
- Clients do not object to AI when they understand it is being used to serve them better. Transparency about AI use actually builds trust.
Frequently Asked Questions
Is AI regulated in insurance?
Yes. The NAIC has issued an AI Model Bulletin, Colorado has passed AI-specific insurance legislation (SB 21-169), and multiple states are developing AI governance frameworks. Key requirements include audit trails, human oversight, fairness testing, and transparency.
How do insurance agents use AI responsibly?
Adopt AI tools with audit trails, human review modes, and documented governance. Review AI outputs regularly, test for fairness, and maintain compliance documentation. Use platforms like unLocked CRM that build compliance features (DNC checks, Review mode, custom instructions) into the AI tools.