US AI Policy
The United States has taken a different approach to AI regulation than the EU, relying on executive orders, voluntary frameworks, and existing sector-specific laws rather than comprehensive legislation.
Federal Framework
The US approach to AI regulation is characterized by voluntary frameworks and sector-specific application of existing laws:
| Framework | Year | Key Elements |
|---|---|---|
| Executive Order on AI | 2023 | Safety testing for powerful models, AI watermarking, privacy protections, equity and civil rights, worker protections |
| NIST AI RMF | 2023 | Voluntary risk management framework with four core functions: Govern, Map, Measure, Manage |
| AI Bill of Rights | 2022 | Non-binding principles: safe systems, algorithmic discrimination protection, data privacy, notice/explanation, human alternatives |
| OMB AI Guidance | 2024 | Requirements for federal agencies using AI: risk management, testing, transparency, human oversight |
NIST AI Risk Management Framework
-
Govern
Establish organizational policies, processes, and structures for managing AI risk. Define roles, responsibilities, and accountability.
-
Map
Identify and document the context in which AI systems operate, including intended uses, stakeholders, potential impacts, and risk factors.
-
Measure
Assess and track identified risks using quantitative and qualitative methods. Monitor for bias, security vulnerabilities, and performance degradation.
-
Manage
Prioritize and respond to identified risks. Implement mitigations, document decisions, and communicate with stakeholders.
Sector-Specific Regulation
In the US, existing laws are being applied and adapted to cover AI:
Financial Services
Fair lending laws (ECOA, FHA), SEC oversight of algorithmic trading, OCC guidance on model risk management, CFPB enforcement on discriminatory AI.
Healthcare
FDA regulation of AI as medical devices, HIPAA privacy protections for health data, CMS rules on AI in Medicare/Medicaid decisions.
Employment
EEOC guidance on AI and anti-discrimination, Title VII applicability to algorithmic hiring, state laws requiring bias audits (NYC Local Law 144).
Transportation
NHTSA regulation of autonomous vehicles, FAA rules for autonomous drones, DOT guidance on AI in safety-critical systems.
State-Level AI Laws
| State | Law/Regulation | Focus |
|---|---|---|
| New York City | Local Law 144 (2023) | Requires bias audits of automated employment decision tools |
| Colorado | AI Act (2024) | Requires developers and deployers to manage risks of high-risk AI systems |
| California | Multiple bills | AI transparency, deepfake disclosure, automated decision systems |
| Illinois | AI Video Interview Act | Requires consent and disclosure for AI analysis in video job interviews |
| Texas | HB 2060 (2025) | Establishes AI advisory council, requires state agency AI use disclosure |
Lilly Tech Systems