US Executive Order on AI: What Organizations Need to Know

In October 2023, President Biden signed Executive Order 14110, establishing the most comprehensive federal action on AI in US history. This landmark order sets new standards for AI safety, security, and trustworthiness across the federal government and beyond.

What is the Executive Order on AI?

Executive Order 14110, titled “Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence,” represents a whole-of-government approach to AI governance. It directs federal agencies to:

• Establish new safety and security standards
• Protect Americans’ privacy
• Advance equity and civil rights
• Support workers affected by AI
• Promote innovation and competition
• Advance American leadership globally

Key Requirements by Area

AI Safety and Security

The order establishes significant safety requirements for powerful AI systems:

Reporting Requirements

• Developers of dual-use foundation models must notify the government
• Companies training models above certain compute thresholds must report
• Results of red-team safety tests must be shared with government

Safety Standards

• NIST must develop standards for red-teaming AI systems
• Guidelines for secure AI development practices
• Standards for AI-generated content authentication (watermarking)

Protecting Privacy

The order addresses AI’s impact on personal privacy:

• Strengthen privacy-preserving techniques in AI
• Evaluate federal agencies’ data collection practices
• Develop guidelines for AI use in surveillance
• Support research on privacy-preserving AI

Advancing Equity and Civil Rights

To prevent AI-enabled discrimination:

• Guidance on algorithmic discrimination in hiring
• Best practices for AI in the justice system
• Standards for AI use in healthcare
• Protections for tenants from AI screening

Supporting Workers

Recognizing AI’s workforce impact:

• Study AI’s labor market effects
• Develop principles for employee well-being
• Support workers displaced by AI
• Guidelines for AI in workplace monitoring

Promoting Innovation

To maintain US competitiveness:

• Streamline visa processes for AI talent
• Expand AI research funding
• Promote small business AI adoption
• Open access to government AI resources

Agency-Specific Mandates

The EO assigns specific responsibilities to federal agencies:

Agency	Key Responsibilities
Commerce/NIST	Safety standards, testing guidelines, watermarking
DHS	AI security in critical infrastructure
DOE	AI risks in nuclear/biological domains
HHS	AI safety in healthcare and drug development
DOL	Worker protections and labor impact
DOJ	Civil rights enforcement and AI in justice
FTC	Consumer protection and competition
OMB	Federal AI procurement and use

Compliance Timeline

The EO established aggressive timelines:

Deadline	Requirement
90 days	Reporting requirements for dual-use models
150 days	NIST AI safety guidelines
180 days	Agency AI use case inventories
240 days	Procurement guidelines
365 days	Comprehensive safety standards

Impact on Private Sector

While the EO primarily directs federal agencies, it significantly affects private companies:

Direct Requirements

• Reporting: Companies developing large AI models must report to government
• Red-teaming: Safety testing results must be shared
• Watermarking: Content authentication for AI-generated media

Indirect Effects

• Federal contractors: Must comply with new AI procurement rules
• Healthcare: AI systems face new safety requirements
• Financial services: Enhanced algorithmic discrimination scrutiny
• Employment: New guidelines on AI in hiring

Relationship to Other Frameworks

The EO complements existing AI governance:

Framework	Relationship
NIST AI RMF	EO references and builds on RMF
State Laws	EO encourages consistent federal approach
EU AI Act	Creates basis for international alignment
Sector Rules	Adds to existing industry regulations

Implementation Status

Key developments since the EO:

• NIST: Released initial safety guidelines and testing frameworks
• OMB: Published federal AI use guidance
• Commerce: Established AI Safety Institute
• DHS: Released critical infrastructure guidance
• Agencies: Completed AI use case inventories

Preparing for Compliance

Organizations should take these steps:

Assessment Phase

1. Inventory AI systems: Document all AI applications
2. Evaluate scope: Determine which EO provisions apply
3. Gap analysis: Compare current practices to requirements
4. Risk assessment: Identify high-priority compliance areas

Implementation Phase

1. Safety testing: Implement red-teaming for high-risk systems
2. Documentation: Prepare required reports and disclosures
3. Governance: Establish oversight structures
4. Training: Educate staff on new requirements

Monitoring Phase

1. Track guidance: Monitor agency rulemaking
2. Update practices: Adapt to evolving requirements
3. Engage stakeholders: Participate in public comment periods
4. Audit compliance: Regular self-assessment

Enforcement Considerations

While the EO itself doesn’t create new legal penalties, enforcement flows through:

• Existing agency authorities (FTC, DOJ, etc.)
• Federal procurement requirements
• Sector-specific regulations
• Future legislation building on EO

How Metrica.uno Helps

Metrica.uno supports EO compliance by:

• Tracking relevant requirements for your AI systems
• Mapping systems to agency-specific mandates
• Generating documentation for federal reporting
• Assessing alignment with NIST guidelines
• Monitoring evolving implementation guidance

Start your assessment to understand your EO compliance posture.