U.S. Executive Order 14110 on Safe, Secure, and Trustworthy AI
Biden's AI executive order is the most aggressive federal AI action yet. Foundation model developers now have to report to the government when they start training runs above certain compute thresholds, share safety test results, and follow red-teaming standards NIST is developing. If you are building large language models, this changes everything. If you are deploying them, expect procurement requirements and sectoral guidance coming fast.
Reviewed for accuracy by Kodi C.
On 30 October 2023 the White House issued Executive Order 14110, the most expansive U.S. federal directive on artificial intelligence to date. The order requires developers of dual-use foundation models share safety test results with the federal government under the Defense Production Act, tasks NIST with advancing red-teaming and watermarking standards, and instructs DHS and other agencies to issue sectoral guidance.
It also orders the development of privacy-preserving techniques, directs agencies to combat algorithmic discrimination, and sets timelines for guidance on critical infrastructure, healthcare, education, and labor impacts. Organizations building or deploying advanced models should track forthcoming reporting obligations, security benchmarks, and procurement conditions shaped by the order.
Defense Production Act Reporting
The order invokes Defense Production Act authorities requiring developers of dual-use foundation models to notify the federal government when starting training runs exceeding specified computational thresholds. Notifications must include information about model capabilities, safety assessments, and red-team findings. This is the first use of DPA reporting requirements for AI development activities.
The computational threshold targets models requiring significant computing resources, initially focused on systems training using approximately 10^26 floating-point operations. Organizations operating large-scale training infrastructure should assess whether planned development activities trigger reporting obligations under forthcoming Commerce Department implementing regulations.
NIST Standards Development
The order directs NIST to develop standards, guidelines, and good practices for AI safety and security. Priority areas include red-teaming methodologies enabling systematic identification of model vulnerabilities, watermarking techniques for AI-generated content authentication, and testing protocols for evaluating safety characteristics.
NIST must establish consensus processes engaging industry, civil society, and academic teams. Standards development timelines create urgency for organizations seeking to influence outcomes. Active participation in NIST processes positions organizations favorably as requirements translate into procurement conditions and regulatory expectations.
Sectoral Guidance Requirements
Cabinet departments receive directives to issue guidance addressing AI risks within their jurisdictions. HHS must address healthcare AI applications including clinical decision support and administrative automation. DOE focuses on critical infrastructure protection. DHS addresses transportation, telecommunications, and cybersecurity applications.
Education, labor, housing, and financial services agencies also receive specific mandates. Organizations operating in regulated industries should monitor sectoral guidance development and engage agency consultation processes shaping setup.
Federal Procurement Implications
The order directs OMB to establish requirements for AI risk management in federal acquisition. Contractors providing AI systems to government agencies will face obligations around testing, documentation, and ongoing monitoring. These requirements affect existing contracts as well as new procurements.
Organizations selling AI products or services to the federal government should prepare for improved scrutiny of AI governance practices. Procurement requirements frequently establish de facto industry standards as commercial customers adopt government-driven expectations.
Privacy and Civil Rights Protections
The order addresses privacy concerns through directives on privacy-preserving machine learning techniques and restrictions on AI-enabled surveillance. Agencies must evaluate AI systems for potential discrimination impacts before deployment. Civil rights considerations extend to housing, employment, credit, and public services applications.
Impact assessment requirements apply to federal agencies and influence expectations for private sector organizations through procurement conditions and sectoral guidance. If you are affected, develop assessment methodologies evaluating disparate impact risks across protected characteristics.
Workforce and Economic Impacts
The order acknowledges AI impacts on employment, directing studies on automation effects and workforce development needs. Labor-related guidance addresses surveillance in the workplace, algorithmic management, and worker notification requirements. These provisions signal forthcoming federal attention to employment-related AI applications.
Schedule and deadlines
Agency actions proceed under aggressive timelines, with many deliverables required within 90-180 days. If you are affected, monitor agency activities and engage emerging consultation opportunities. The compressed setup schedule creates compliance urgency for organizations subject to forthcoming requirements.
- Executive Order text provides detailed directives, timelines, and agency roles.
- White House fact sheet summarizes key requirements for model developers and federal agencies.
Continue in the Policy pillar
Return to the hub for curated research and deep-dive guides.
Latest guides
-
AI Policy Implementation Guide
Coordinate governance, safety, and reporting programmes that meet EU Artificial Intelligence Act timelines and U.S. National AI Initiative Act mandates while sustaining product…
-
Digital Markets Compliance Guide
Implement EU Digital Markets Act, EU Digital Services Act, UK Digital Markets, Competition and Consumers Act, and U.S. Sherman Act requirements with cross-functional operating…
-
Semiconductor Industrial Strategy Policy Guide
Coordinate CHIPS and Science Act, EU Chips Act, and Defense Production Act programmes with capital planning, compliance, and supplier readiness.
Coverage intelligence
- Published
- Coverage pillar
- Policy
- Source credibility
- 71/100 — medium confidence
- Topics
- Executive Order · Model Safety · Regulation · United States
- Sources cited
- 2 sources (iso.org, crsreports.congress.gov)
- Reading time
- 5 min
References
- Industry Standards and Best Practices — International Organization for Standardization
- Congressional Research Service Analysis
Comments
Community
We publish only high-quality, respectful contributions. Every submission is reviewed for clarity, sourcing, and safety before it appears here.
No approved comments yet. Add the first perspective.