Joe Braidwood
Joe Braidwood
CEO, GLACIS
35 min read

Executive Summary

The United States lacks comprehensive federal AI legislation, creating a rapidly evolving patchwork of state-level regulations. As of December 2025, Colorado has enacted the most comprehensive AI law (effective June 2026), while California, Illinois, Texas, New York, and over 30 other states have introduced or are actively considering AI-specific legislation.

State AI laws generally address three categories: (1) algorithmic discrimination in high-stakes decisions, (2) automated decision-making transparency and consumer rights, and (3) sector-specific AI use in employment, healthcare, insurance, and financial services. Many laws leverage existing consumer protection frameworks rather than creating entirely new regulatory structures.

Key compliance insight: Organizations operating nationally should adopt a "highest common denominator" approach—implementing controls that satisfy the strictest applicable requirements. Frameworks like NIST AI RMF provide structured approaches that map to multiple state requirements, reducing duplicative compliance efforts.

1
Comprehensive AI Law Enacted
30+
States with AI Bills
Jun 2026
Colorado AI Act Effective
$20K
Max Penalty per Violation

In This Guide

The US AI Regulatory Landscape

Unlike the European Union, which enacted a comprehensive AI Act covering all member states, the United States has taken a fragmented approach to AI regulation. In the absence of federal legislation, individual states have begun enacting their own AI laws—creating a complex compliance landscape for organizations operating across state lines.

Why States Are Acting

Several factors are driving state-level AI regulation:

Types of State AI Laws

State AI legislation generally falls into several categories:

Categories of State AI Legislation

Category Focus Example States
Comprehensive AI Laws Broad regulation of high-risk AI systems across multiple domains Colorado (enacted), California (pending)
Employment AI AI in hiring, promotion, termination decisions Illinois (AIPLA), New York (Local Law 144), Maryland
Biometric AI Facial recognition, voice recognition, biometric data Illinois (BIPA), Texas, Washington
Privacy + AI Automated decision-making provisions in privacy laws California (CCPA/CPRA), Virginia (VCDPA), Connecticut (CTDPA)
Healthcare AI AI in clinical decisions, insurance, care management California (pending), New York (proposed)
Government AI AI use by state and local government agencies California, Washington, multiple states

Colorado: The First Comprehensive State AI Law

Colorado's Artificial Intelligence Act (SB 24-205), signed May 17, 2024 and effective June 30, 2026, is the first comprehensive US state law regulating AI systems. It establishes obligations for both "developers" (those who build AI) and "deployers" (those who use AI in consequential decisions).

Enacted Effective June 30, 2026

Colorado AI Act Key Points

  • Scope: High-risk AI in employment, housing, credit, healthcare, education, insurance, government services, legal services
  • Standard: "Reasonable care" to prevent algorithmic discrimination
  • Safe Harbor: NIST AI RMF or ISO 42001 compliance creates rebuttable presumption
  • Penalties: Up to $20,000 per violation (Consumer Protection Act)
  • Enforcement: Attorney General only (no private right of action)

Developer Requirements

Developers of high-risk AI systems must:

Deployer Requirements

Deployers of high-risk AI systems must:

For comprehensive coverage, see our Colorado AI Act Complete Compliance Guide.

California: The Privacy Leader Expands to AI

California leads US states in data privacy regulation, and its frameworks increasingly address AI. While California hasn't enacted a comprehensive AI law equivalent to Colorado's, multiple overlapping regulations affect AI deployment:

California Consumer Privacy Act (CCPA/CPRA)

Enacted Effective January 1, 2023 (CPRA amendments)

CCPA/CPRA AI Provisions

  • Profiling opt-out: Consumers can opt out of automated decision-making
  • Access rights: Consumers can access information about automated decisions
  • Risk assessments: Required for processing posing significant risk (including profiling)
  • Penalties: $2,500-$7,500 per intentional violation

California Automated Decision-Making Technology (ADMT) Regulations

The California Privacy Protection Agency (CPPA) is developing comprehensive ADMT regulations that will significantly expand AI requirements. Key provisions under consideration:

These regulations are expected to take effect in 2026 or 2027 following the formal rulemaking process.

California Pending AI Legislation

California's legislature has considered multiple AI bills, including proposals modeled on the EU AI Act:

Illinois: Biometrics and Employment AI Pioneer

Illinois has been at the forefront of regulating specific AI applications, particularly biometric data and employment decisions:

Illinois Biometric Information Privacy Act (BIPA)

Enacted Effective 2008

BIPA Requirements

  • Scope: Fingerprints, face geometry, iris scans, voice prints, hand geometry
  • Notice & consent: Written consent required before collection
  • Private right of action: Individuals can sue directly
  • Penalties: $1,000 per negligent violation; $5,000 per intentional violation

BIPA has generated significant litigation against AI companies using facial recognition technology, with settlements reaching hundreds of millions of dollars. The law effectively prohibits most commercial facial recognition uses without explicit consent.

Illinois Artificial Intelligence Video Interview Act (AIVIA)

Enacted Effective January 1, 2020

AIVIA Requirements

Employers using AI to analyze video interviews must: (1) notify applicants that AI will be used; (2) explain how the AI works and what characteristics it evaluates; (3) obtain applicant consent before the interview; (4) limit who can view the video; (5) delete videos upon applicant request.

Illinois Employment AI Legislation

Illinois continues to expand employment AI regulation:

Texas: Emerging AI Regulation

Texas, with its large technology sector and business-friendly reputation, has taken a measured approach to AI regulation while addressing specific concerns:

Texas Capture or Use of Biometric Identifier Act (CUBI)

Enacted Effective 2009

Texas CUBI

Requires notice and consent before capturing biometric identifiers for commercial purposes. Unlike Illinois BIPA, Texas does not provide a private right of action—enforcement is through the Attorney General. Penalties up to $25,000 per violation.

Texas Data Privacy and Security Act (TDPSA)

Effective July 1, 2024, the TDPSA includes provisions affecting AI:

Texas AI Advisory Council

Texas established an AI Advisory Council to study AI issues and recommend legislation. Areas under consideration include:

New York: Local and State AI Regulation

New York presents a complex regulatory landscape with both city-level and state-level AI requirements:

New York City Local Law 144 (Automated Employment Decision Tools)

Enacted Effective July 5, 2023

NYC Local Law 144

  • Scope: Automated employment decision tools (AEDTs) used in NYC hiring/promotion
  • Bias audit: Annual independent audit for disparate impact by race, ethnicity, sex
  • Publication: Audit summary must be publicly posted
  • Notice: Candidates must be notified at least 10 days before AEDT use
  • Penalties: $500 first violation; $500-$1,500 subsequent violations per day

New York State AI Legislation

New York State has multiple pending and enacted AI-related laws:

Other State AI Laws and Pending Legislation

Beyond the major states covered above, AI regulation is advancing across the country:

States with Privacy Laws Including AI Provisions

Virginia (VCDPA)

Enacted

Effective January 1, 2023

  • • Profiling opt-out rights
  • • Data protection assessments for profiling
  • • No private right of action

Connecticut (CTDPA)

Enacted

Effective July 1, 2023

  • • Profiling opt-out for legal/significant decisions
  • • Data protection assessments required
  • • 60-day cure period

Utah (UCPA)

Enacted

Effective December 31, 2023

  • • Consumer access to profiling information
  • • More limited than other state laws
  • • AG enforcement only

Montana (MCDPA)

Enacted

Effective October 1, 2024

  • • Profiling opt-out rights
  • • Data protection assessments
  • • 60-day cure period

Oregon (OCPA)

Enacted

Effective July 1, 2024

  • • Profiling opt-out for automated decisions
  • • Data protection assessments
  • • Cure period through 2026

Delaware (DPDPA)

Enacted

Effective January 1, 2025

  • • Profiling opt-out rights
  • • No revenue threshold
  • • Broad applicability

States with Biometric/Facial Recognition Laws

State Law Private Action Key Requirements
Illinois BIPA Yes Most stringent; written consent required
Texas CUBI No Notice and consent; AG enforcement
Washington HB 1493 No Notice required; enrollment consent
Arkansas PIPA No Notice and consent requirements
Maryland SB 169 No Facial recognition restrictions in employment

States with Government AI Restrictions

Several states have enacted or proposed restrictions on government use of AI:

State AI Law Comparison Matrix

This matrix provides a high-level comparison of key AI regulatory requirements across major states:

Requirement Colorado California Illinois New York Texas
Comprehensive AI Law ✓ Enacted Pending Partial Pending
Employment AI CPRA ✓ AIVIA ✓ LL144
Biometric AI CCPA ✓ BIPA Pending ✓ CUBI
Impact Assessments ✓ Required ✓ CPRA ✓ LL144 ✓ TDPSA
Consumer Opt-Out Limited
Private Right of Action No Limited Yes (BIPA) No No
Safe Harbor (Frameworks) ✓ NIST/ISO

Federal Context: No Comprehensive Law Yet

Understanding the federal landscape helps contextualize state action:

Why No Federal AI Law?

Despite bipartisan interest in AI regulation, federal legislation has stalled due to:

Existing Federal AI-Related Laws

While no comprehensive AI law exists, several federal laws affect AI deployment:

Federal Agency Guidance

Federal agencies have issued AI guidance within their regulatory domains:

NIST AI Risk Management Framework

The NIST AI RMF, released January 2023, provides a voluntary framework that multiple state laws reference. Colorado explicitly provides a safe harbor for organizations following NIST AI RMF or ISO 42001.

Multi-State Compliance Strategy

Organizations operating across multiple states need a strategic approach to managing divergent requirements:

Highest Common Denominator Approach

Rather than maintaining separate compliance programs for each state, implement controls satisfying the strictest applicable requirements:

GLACIS Framework

Multi-State AI Compliance

1

Adopt NIST AI RMF

Implement NIST AI Risk Management Framework as baseline. Provides Colorado safe harbor and maps to most state requirements. Document implementation across all four functions: Govern, Map, Measure, Manage.

2

Implement Comprehensive Impact Assessments

Create impact assessment templates that satisfy Colorado, California CPRA, NYC LL144, and pending state requirements. Include bias testing, discrimination risk analysis, and consumer rights documentation.

3

Build Unified Consumer Rights Infrastructure

Implement consumer-facing capabilities: opt-out mechanisms, explanation rights, data correction, appeal processes with human review. Design once, deploy across all states.

4

Document for Multiple Regulators

Maintain documentation that can be adapted for any state regulator: risk management policies, bias testing results, training records, incident response procedures. Use standardized formats (model cards, dataset cards).

5

Monitor Regulatory Evolution

Establish processes to track new state legislation, regulatory guidance, and enforcement actions. Update compliance programs proactively rather than reactively. Subscribe to AG office updates and industry associations.

Sector-Specific Considerations

Certain industries face additional state-specific requirements:

Healthcare AI

Employment AI

Financial Services AI

Frequently Asked Questions

Which state has the strictest AI law?

Colorado currently has the most comprehensive state AI law with broad coverage of high-risk AI systems. However, Illinois BIPA is the strictest for biometric AI due to its private right of action and significant damages. For employment AI, NYC Local Law 144 sets rigorous bias audit requirements. California's pending ADMT regulations may eventually surpass Colorado in some respects.

Do state AI laws apply to companies headquartered elsewhere?

Yes. State AI laws typically apply based on where consumers are located, not where companies are headquartered. If you serve Colorado residents, make decisions affecting Illinois employees, or deploy AI impacting NYC job candidates, you must comply with those jurisdictions' laws regardless of your company's location.

Will federal AI law preempt state laws?

Uncertain. If comprehensive federal AI legislation passes, it may or may not preempt state laws depending on the law's language. Historically, federal privacy laws (like HIPAA and FCRA) have included limited preemption, allowing states to enact more protective requirements. Current state AI laws generally don't conflict with federal requirements—they fill gaps in federal coverage.

How do I know which state laws apply to my AI system?

Consider: (1) Where are the people affected by your AI decisions located? (2) What type of AI application is it (employment, healthcare, credit, etc.)? (3) What data does it process (biometric, personal information)? (4) Who deploys the system (government, private sector)? Most organizations operating nationally should assume the strictest applicable requirements apply.

What's the difference between a developer and deployer under state AI laws?

Developers create or substantially modify AI systems (model providers, algorithm developers). Deployers use AI systems to make decisions affecting consumers (employers using hiring AI, lenders using credit scoring). An organization can be both if they build and use their own AI. Each role has distinct compliance obligations under laws like the Colorado AI Act.

Do small businesses need to comply with state AI laws?

It depends on the law. Some states (like California and Virginia) have revenue or data volume thresholds. Colorado AI Act applies based on high-risk AI use, not company size. NYC LL144 applies to any employer using AEDTs in NYC hiring. Illinois BIPA has no size exemption. Check specific law thresholds, but assume requirements apply if you're using high-risk AI.

Key Takeaways

  • Colorado leads: First comprehensive state AI law, effective June 2026 with reasonable care standard
  • Patchwork is growing: 30+ states have AI bills; major states have enacted targeted laws
  • NIST AI RMF provides safe harbor: Colorado explicitly recognizes framework compliance
  • Illinois BIPA is highest risk: Private right of action creates significant litigation exposure
  • National companies need unified approach: Implement highest common denominator controls
  • More regulation coming: California ADMT, healthcare AI bills, and new state laws in 2026

Navigate Multi-State AI Compliance

Our Evidence Pack demonstrates compliance across multiple state frameworks—Colorado AI Act, CCPA/CPRA, and beyond. One investment, multi-jurisdictional coverage.

Build Your Evidence Pack

Related Guides