A complete state-by-state matrix of AI disclosure laws affecting insurance agents in 2026 — covering chatbots, voice AI, automated underwriting.
TL;DR
As of January 2026, 14 US states have enacted AI disclosure laws that affect insurance agent operations. California (AB 2013, SB 942), Colorado (SB 24-205), Utah (SB 149), and Texas (HB 2060) have the strictest requirements. Common obligations: disclose AI when consumers interact with chatbots/voice agents, label AI-generated content, and provide opt-out paths to human agents. Penalties range from $2,500 (Utah) to $25,000 per violation (California).
Key Takeaways
- 14 states have enacted AI disclosure laws affecting insurance as of Jan 2026
- California (AB 2013, SB 942) and Colorado (SB 24-205) have the strictest requirements
- Most laws require: disclosure of AI use, opt-out to human, content labeling
- Utah requires real-time disclosure when consumers ask if they're talking to AI
- Penalties range from $2,500 to $25,000 per violation
State-by-State AI Disclosure Laws for Insurance Agents
Updated January 15, 2026. Reflects laws effective on or before January 1, 2026.
Why This Matters Now
The state-level AI legal landscape exploded in 2024-2025. Insurance agents using AI tools — chatbots, voice AI, autonomous outbound, AI-generated emails — are increasingly subject to disclosure obligations that vary state-by-state.
Failure to comply can trigger DOI complaints, attorney general enforcement, and E&O exposure.
The Four Core Compliance Obligations
Across the 14 states with active laws, the same four obligations recur:
- Disclosure of AI use — tell consumers when they're interacting with AI
- Opt-out to a human — provide a path to a human agent
- Content labeling — mark AI-generated content (text, images, voice)
- Documentation — retain logs of AI interactions and disclosures
State-by-State Highlights
California (AB 2013, SB 942) — Strictest
- AB 2013: Requires disclosure of training data sources for generative AI
- SB 942: Requires visible + metadata labels on AI-generated content
- Penalty: up to $25,000/violation
Colorado (SB 24-205) — Most Comprehensive
- Effective Feb 1, 2026 — covers "high-risk AI" in consumer-facing decisions
- Requires algorithmic impact assessments, bias testing, consumer notices
- Penalty: up to $20,000/violation
Utah (SB 149) — Real-Time Disclosure
- Requires AI to disclose itself when asked ("Am I talking to a bot?")
- Applies to chatbots and voice AI
- Penalty: $2,500-$10,000/violation
Texas (HB 2060)
- Established the Texas AI Advisory Council; rules expanding 2026
- Sector-specific rules for regulated industries (insurance included)
Connecticut (SB 1103)
- Requires consumer disclosure when generative AI is used in consumer services
- Penalty: $1,000-$10,000
Tennessee (ELVIS Act)
- Specifically targets voice/likeness cloning
- Critical for agents using voice cloning for callbacks
How to Stay Compliant
- Default to disclosure — proactively disclose AI use in chat, voice, and email signatures
- Maintain a human-handoff path in every AI interaction
- Label AI content — use platform-native labels (unLocked auto-labels generated content)
- Keep audit logs — minimum 24 months recommended
- Update your privacy notice to describe AI use and consumer rights
How unLocked Handles This
The unLocked compliance layer auto-applies state-appropriate disclosure language based on the consumer's state, retains 7-year interaction logs, labels AI-generated content with both visible and metadata markers, and provides one-click human handoff in every channel.
State AI Disclosure Laws Affecting Insurance (2026)
State — Law — Trigger — Penalty
California — AB 2013 + SB 942 — Generative AI use; chatbot interactions — Up to $25,000/violation
Colorado — SB 24-205 — High-risk AI in consumer decisions — Up to $20,000/violation
Utah — SB 149 — Real-time AI disclosure on request — $2,500-$10,000/violation
Texas — HB 2060 — AI in govt/regulated sectors — Civil penalties + injunctive relief
Illinois — HB 3773 (employment focus, partial) — AI in screening (extends to producers hiring) — Civil penalties
New York — AI bias audit (NYC Local Law 144) — Automated employment decisions — $500-$1,500/violation
Florida — Bot disclosure (election-adjacent) — Limited insurance impact — Civil penalties
Connecticut — SB 1103 — Generative AI in consumer-facing services — $1,000-$10,000
Tennessee — ELVIS Act — Voice/likeness cloning — Civil + criminal
Virginia — HB 2154 (pending) — Consumer AI disclosure — TBD
Washington — AI Task Force Rules — Adverse decisions, healthcare adjacency — $5,000-$15,000
Oregon — SB 1571 — Synthetic media disclosure — Civil penalties
Massachusetts — S.31 (pending) — Consumer AI disclosure — TBD
Michigan — HB 5141 (pending) — Voice AI disclosure — TBD
Frequently Asked Questions
Which states have AI disclosure laws affecting insurance agents in 2026?
As of January 2026, 14 US states have enacted AI disclosure laws affecting insurance: California, Colorado, Utah, Texas, Illinois, New York, Florida, Connecticut, Tennessee, Virginia (pending), Washington, Oregon, Massachusetts (pending), and Michigan (pending).
Do I have to tell prospects when I'm using AI to call them?
In most states, yes — if asked. Utah's SB 149 requires real-time disclosure when a consumer asks if they're interacting with AI. California (SB 942) requires disclosure for generative AI content. Best practice: always disclose AI use upfront in voice and chat interactions, and document consent.
What is California SB 942?
California SB 942 (the California AI Transparency Act) requires providers of generative AI systems to label AI-generated content with both visible and embedded (metadata) disclosures. It applies to insurance agents using AI for marketing content, voice scripts, or generated communications.
What is the penalty for violating state AI disclosure laws?
Penalties range from $2,500 per violation (Utah SB 149) to $25,000 per violation (California AB 2013/SB 942). Colorado SB 24-205 imposes up to $20,000/violation, plus civil enforcement by the state attorney general.