Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence
The order establishes a sweeping, whole-of-government framework for governing artificial intelligence, requiring large AI model developers to report training activities to the federal government under the Defense Production Act, directing agencies to create safety standards, and coordinating protections for workers, civil rights, consumers, and privacy across virtually every sector of the economy.
It is the first executive order to invoke Defense Production Act and IEEPA authorities to impose direct reporting obligations on AI companies, and it creates a White House AI Council to oversee implementation across more than a dozen federal departments and agencies.
What this order does
What it orders
The order directs a comprehensive, whole-of-government AI governance framework covering safety standards, security, civil rights, workers, consumers, privacy, and federal AI adoption. Using Defense Production Act authority, it requires within 90 days that companies developing the largest AI models report training activities, model weights, and red-team test results to the federal government. It also directs the Secretary of Commerce to propose regulations requiring U.S. cloud computing providers to report when foreign persons use their infrastructure to train powerful AI models — drawing on IEEPA authority tied to an existing cybersecurity national emergency — and establishes a White House AI Council to coordinate implementation across agencies. Dozens of additional directives cover AI safety standards at NIST, biosecurity screening for synthetic biology, synthetic-content watermarking, AI-talent immigration pathways, anti-discrimination in hiring and housing, and AI governance within healthcare, transportation, and education.
Most directives require agencies to study, develop guidance, or propose regulations rather than imposing immediately binding rules on the public. AI used in national security systems is explicitly excluded from most provisions, with governance of that domain reserved for a separately directed National Security Memorandum. Implementation is subject to available appropriations, and the order creates no enforceable legal rights against the government.
Who it affects
Developers of large AI models and computing clusters, U.S. cloud computing providers and their foreign resellers, federal agencies across virtually every sector, the federal workforce using AI tools, businesses in healthcare, financial services, transportation, and education, and workers whose jobs may be affected by AI deployment.
Why it matters
For the first time, companies training the most powerful AI models must disclose their activities to the federal government. Cloud providers must track and report foreign users training large AI models. Every major federal agency gains new governance requirements shaping how AI is used in government-delivered services, from benefits administration to criminal justice.
What must happen and when
How the order is supposed to work
The White House AI Council, chaired by the Deputy Chief of Staff for Policy, coordinates implementation. OMB issues binding guidance to agencies within 150 days; agencies then designate Chief AI Officers within 60 days of that guidance. The Defense Production Act reporting requirement for AI developers flows through the Secretary of Commerce, who defines qualifying technical thresholds. IEEPA authority is delegated to Commerce to promulgate cloud-computing oversight regulations. Dozens of interim agency reports cascade into successive actions, creating a layered implementation arc spanning 30 to 540 days, with no single sunset provision.
Actions and deadlines
- Technology Modernization Board considers prioritizing AI project funding for at least one year
- Secretary of Transportation directs NETT Council to assess AI needs in transportation and establish a Cross-Modal Executive Working Group
- Secretary of Labor publishes a request for information on AI and STEM occupations for the Schedule A list
- Director of OSTP and Director of OMB identify priority AI talent areas and accelerated hiring pathways for the federal government
- Convene the AI and Technology Talent Task Force within the Executive Office of the President
- Director of OMB convenes interagency council to coordinate federal AI development and use
- Director of OPM conducts evidence-based review on direct-hire authority for AI and data-science roles
- Secretary of Commerce requires large AI model developers to report training activities, model weights, and red-team results to the federal government
- Secretary of Commerce proposes regulations requiring U.S. IaaS providers to report when foreign persons conduct large AI training runs on their infrastructure
- NSF Director launches pilot program implementing the National AI Research Resource
- HHS Secretary establishes an HHS AI Task Force in consultation with Defense and Veterans Affairs
- Attorney General convenes heads of federal civil rights offices to address AI-related discrimination
- Administrator of General Services develops and issues a framework for prioritizing AI offerings in FedRAMP
- Secretary of Defense contracts National Academies to study AI-related biosecurity risks including generative AI trained on biological data
- DHS Secretary develops and publishes informational resources for attracting AI talent to the United States
- Director of OMB issues guidance to agencies on federal AI use, risk management, and designation of Chief AI Officers
- Secretary of Treasury issues public report on best practices for financial institutions to manage AI cybersecurity risks
- Secretary of Commerce proposes regulations requiring foreign IaaS resellers to verify identity of foreign account holders
- Secretary of Homeland Security incorporates NIST AI Risk Management Framework into critical infrastructure safety and security guidelines
- Secretary of Defense and Secretary of Homeland Security each develop and complete AI pilot projects for cyber vulnerability discovery
- Secretary of Homeland Security evaluates AI's potential to enable CBRN threats and submits report to the President
- Director of OSTP establishes a framework for synthetic nucleic acid procurement screening to reduce biosecurity risks
- Secretary of Labor develops and publishes workplace AI principles and best practices for employers
- Secretary of Commerce submits report identifying standards, tools, and methods for authenticating and labeling synthetic AI-generated content
- NIST Director establishes guidelines and best practices for safe and secure AI development, including generative AI companion resources
- Secretary of Energy develops AI model evaluation tools and testbeds capable of assessing nuclear, biological, chemical, and infrastructure threats
- Secretary of Commerce solicits public input and submits report on risks and benefits of dual-use foundation models with widely available weights
- OSTP and National Security Advisor oversee development of a proposed National Security Memorandum on AI governance for military and intelligence use
- Attorney General submits report to the President on AI use in the criminal justice system, including sentencing, policing, and forensics
- HHS AI Task Force develops a strategic plan for responsible AI deployment in health and human services
- Secretary of Education develops AI resources, policies, and guidance for the education sector
- NSF Director establishes at least four new National AI Research Institutes