On October 30, 2023, President Biden released an Executive Order (EO) on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence (AI). This landmark EO seeks to advance the safe and secure development and deployment of AI by implementing a society-wide effort across government, the private sector, academia, and civil society to harness “AI for good,” while mitigating its substantial risks.
The EO lays out eight guiding principles and priorities that consider the views of “other agencies, industry, members of academia, civil society, labor unions, international allies and partners, and other relevant organizations” to advance and govern the use of AI. These include:
- Ensure safe and secure AI technology;
- Promote responsible innovation, competition, and collaboration;
- Support American workers;
- Advance equity and civil rights;
- Protect American consumers, patients, passengers, and students;
- Protect privacy and civil liberties;
- Manage the federal government’s use of AI; and
- Strengthen U.S. leadership abroad, safeguarding ways to develop and deploy AI technology responsibly.
These eight principles are detailed in EO Sections 4 through 11. A summary of each section is provided below.
Section 4. Ensuring the Safety and Security of AI Technology
Section 4 focuses on eight key areas: (1) developing guidelines, standards, and best practices for AI safety and security; (2) ensuring safe and reliable AI through industry reporting on AI development and datacenters, and reporting on foreign access to AI infrastructure; (3) managing AI in critical infrastructure and cybersecurity; (4) reducing risks at the intersection of AI and Chemical, Biological, Radiological, and Nuclear (CBRN) threats; (5) reducing the risks posed by synthetic content; (6) soliciting input on dual-use foundation models with widely available model weights; (7) promoting safe release and preventing the malicious use of federal data for AI training; and (8) developing a coordinated executive branch approach to managing AI security risks. The EO directs the Secretary of Commerce and the Department of Homeland Security (DHS) to perform a number of tasks that will further ensure the safe and secure use of AI technology. Some of those tasks include:
- The Secretary of Commerce, acting through the Director of the National Institute of Standards and Technology (NIST), will coordinate with other relevant agencies to establish guidelines and best practices that promote consensus industry standards on developing and deploying AI. The Secretary of Energy is directed to develop AI evaluation tools to identify security risks, including nuclear and energy-security threats.
- The Secretary of Commerce is directed to use the authority of the Defense Production Act (DPA) to require U.S. companies to report on any development of “dual use” AI foundation models, the ownership of resulting model weights, and the results of red-team testing; and to report on the development or possession of any “large scale computing cluster.”
- The Secretary of Commerce is directed to solicit input through a public consultation regarding the risks related to the removal of safeguards with AI models, the risks of actors fine-tuning the foundation models, and the benefits and risks of AI innovation. The EO addresses the security risks of dual-use foundation models, including regulations to curb their use by foreign malicious actors, with a particular focus on models with widely available weights.
- DHS is directed to establish an AI Safety and Security Board as an advisory committee of AI experts to advise on security improvements and incident response related to AI usage in critical infrastructure. To ensure the protection of critical infrastructure, heads of Sector Risk Management Agencies are also directed to assess AI risks related to critical failures, physical attacks, and cyber-attacks.
Section 5. Promoting Innovation and Competition
Section 5 focuses on three key areas: (1) attracting AI talent to the U.S.; (2) promoting innovation; and (3) promoting competition, with key actions required that address health care, energy, and intellectual property. Below are a few notable ways the EO addresses the promotion of innovation and competition:
- The Secretary of the Department of Health and Human Services (HHS) is directed to prioritize grantmaking and cooperative agreement awards, including through the National Institutes of Health, to support responsible AI development and use in the healthcare sector, health data quality, and health equity. The EO also requires the Secretary of Veterans Affairs to host two 3-month nationwide AI Tech Sprint competitions to improve the quality of veterans’ health care and support small businesses’ innovative capacity.
- The EO seeks to promote AI innovation and combat risks to developers. To do so, the Secretary of Commerce and U.S. Patent and Trademark Office are directed to issue guidance on patentability and copyright issues related to AI, and the Secretary of DHS is directed to investigate incidents of intellectual property theft and pursue enforcement.
- The Director of the National Science Foundation (NSF) is directed to launch a pilot program implementing the National AI Research Resource (NAIRR), which will provide training resources to support AI research and development. At least one NSF Regional Innovation Engine and four new National AI Research Institutes must be funded.
Section 6. Supporting Workers
Section 6 focuses on two key areas: (1) advancing the government’s understanding of AI’s implications for workers; and (2) ensuring that AI in the workplace advances employees’ well-being. As highlighted below, the EO made several requests for information and research regarding the impact of AI on the workforce and expressed concerns about deploying AI in the workplace and its effect on workers. The EO also enumerated steps to prioritize diversity in the AI-ready workforce.
- The EO requires several actions to be taken to understand AI’s impact on the workforce, including a report regarding the labor-market effects of AI and a report on the abilities of agencies to support workers displaced by the adoption of AI.
- Focusing on the advancement of employee well-being as AI is deployed in the workplace, the Secretary of Labor is directed to develop and publish principles and best practices that employers can use to mitigate AI’s potential harms to employees’ well-being as well as maximize AI’s potential benefits.
- The EO requires the Secretary of Labor to issue guidance establishing that employers who deploy AI to monitor or augment employees’ work must still compensate employees appropriately for hours worked, as defined under the Fair Labor Standards Act.
- The Administration indicates its interest in fostering diversity within AI-related industries by authorizing the prioritization of resources to support AI-related education and workforce development.
Section 7. Advancing Equity and Civil Rights
Section 7 focuses on three key categories where AI can impact civil rights: (1) the criminal justice system; (2) government benefits and programs; and (3) issues in the broader economy, including hiring, housing, and consumer finance. Specifically, the EO directs and requires various executive agencies to report on how AI may be used in discriminatory ways and to issue guidance on how such potentially discriminatory practices can be mitigated, as highlighted below.
- The Attorney General is directed to evaluate and report on how existing laws address civil rights violations and discrimination stemming from AI usage, and to meet with the heads of federal civil rights offices to discuss comprehensive prevention of AI-related discrimination.
- The Attorney General is also directed to prepare a report by the end of October 2024 that outlines whether and how AI is currently used within the criminal justice system, and identifying best practices for the usage of AI in the criminal justice system.
- The Secretary of HHS and the Secretary of Agriculture are directed will issue guidance concerning the use of AI to maximize program participation by eligible recipients, and outlining when access to and/or decision-making by human reviewers (e.g. as related to benefits determinations) is warranted.
- The Secretary of Labor, Federal Housing and Finance Agency, Consumer Financial Protection Bureau, Secretary of Housing and Urban Development, and the Architectural and Transportation Barriers Compliance Board will evaluate, publish guidance, and/or solicit public comment on various AI-related issues. The overarching objective is to ensure nondiscrimination in the substantive areas that fall within their purview, such as hiring, tenant screening, advertising, and the use of biometric data.
Section 8. Protecting Consumers, Patients, Passengers, and Students
Section 8 focuses on three points: (1) encouraging independent regulatory agencies to consider the full range of authorities to protect American consumers from fraud, discrimination, and threats to privacy, and emphasizing or clarifying requirements and expectations for transparency in AI models; (2) ensuring the safe and responsible deployment and development of AI in healthcare, public health, human-services, transportation, and education sectors; and (3) considering how AI will affect communications networks. The EO requires the following actions from agencies to further the protection of consumers, patients, passengers, and students.
- Independent regulatory agencies will address risks that may arise from the use of AI, including risks to financial stability. The EO directs Independent Regulatory Agencies to clarify responsibilities, conduct due diligence and monitor third-party AI services they use, and to clarify expectations and requirements related to the transparency and explainability of AI models.
- The Secretary of HHS will establish an HHS AI Task Force to create a strategic plan that includes policies and frameworks for the responsible deployment and use of AI in the health and human services sector, as well as establish an AI Safety Program with voluntary federally listed Patient Safety Organizations. Separately, the Secretary of HHS will establish a quality strategy, including an AI assurance policy, advance compliance with federal nondiscrimination laws by providers, and develop a strategy for regulating the use of AI in drug development.
- The Secretary of Transportation will direct the Nontraditional and Emerging Transportation Technology (NETT) Council to assess the need for information, technical assistance and guidance regarding AI in technology. They will support existing and future initiatives that pilot transportation-related applications of AI, evaluate the outcomes of such pilot programs, and establish a new Department of Transportation (DOT) Cross-Modal Executive Working Group to coordinate applicable work. In addition, the Advanced Research Projects Agency-Infrastructure (ARPA-I) will explore and solicit public opinion on transportation-related opportunities and challenges.
- The Secretary of Education will develop resources, policies, and guidance regarding AI, addressing safe, responsible and nondiscriminatory uses of AI in education, as well as their impact on vulnerable and underserved communities. The Secretary of Education should develop an “AI Toolkit” for education leaders, as well as implement recommendations from the Department of Education’s AI and the Future of Teaching and Learning report.
- The Federal Communications Commission (FCC) will consider how AI affects communications networks and consumers, including by examining the potential for AI to improve spectrum management, creating opportunities for sharing spectrum between federal and non-federal spectrum operations, supporting improved network security, resiliency, and interoperability, and conducting efforts to combat unwanted robocalls and robotexts that are facilitated or exacerbated by AI.
Section 9. Protecting Privacy
Section 9 mandates federal government action to mitigate potential threats to privacy posed by AI. In particular, the EO expresses concern regarding “AI’s facilitation of the collection or use of information about individuals, or the making of inferences about individuals.” Accordingly, the EO obligates the federal government to take the following actions with regards to protecting privacy:
- The Office of Management and Budget (OMB) must take steps to identify commercially available information (CAI) procured by federal agencies, defined as “information or data about an individual or group of individuals, including their device or location, that is made available or obtainable and sold, leased, or licensed to the general public or to governmental or non-governmental entities.”
- OMB will evaluate agency standards for the collection, processing, or use of CAI that contains Personally Identifiable Information (PII), and issue a Request for Information to inform potential revisions to such standards by the end of April 2024.
- By the end of October 2024, NIST must create guidelines for agencies to evaluate “differential-privacy-guarantee protections,” defined as protections that allow information about a group to be shared while limiting the leakage of personal information.
- The National Science Foundation (NSF) will promote research, development, and implementation of privacy-enhancing technologies (PETS), including by creating a Research Coordination Network dedicated to advancing privacy research and working with federal agencies to identify opportunities to incorporate PETS (e.g., AI-generated synthetic data) into agency operations.
Section 10. Advancing Federal Government Use of AI
Section 10 provides direction for federal government agency efforts to develop and use AI technology (1) focusing on the provision of government-wide guidance for agency use, management, and procurement of AI; and (2) outlining a series of priorities and initiatives intended to improve and accelerate federal hiring of AI talent and provide AI training to federal employees
Providing Guidance for AI Management
- The Director of OMB will convene and chair an interagency council to coordinate agencies’ development and use of AI. OMB will issue guidance to govern agency AI usage, advance AI innovation, and manage risks posed by the federal government’s use of AI. Notably, this does not apply to the use of AI in national security systems.
- OMB must take steps to ensure that agency procurement of AI systems and services align with the guidance. The OMB AI guidance must include the designation of a Chief AI Officer and AI Governance Board at each federal agency; required minimum risk-management practices for government use of AI; recommendations to agencies regarding a variety of AI governance topics, including external AI testing; reporting requirements; and guidelines governing the federal workforce’s use of generative AI.
- OMB will also develop a framework to prioritize critical and emerging cloud offerings in the Federal Risk and Authorization Management Program (FedRAMP) authorization process. OMB is instructed to begin by prioritizing “generative AI offerings whose primary purpose is to provide large language model-based chat interfaces, code-generation and debugging tools, and associated application programming interfaces, as well as prompt-based image generators.”
- The General Services Administration (GSA), in coordination with OMB, the Department of Defense (DOD), DHS, NASA, and other federal agencies, is directed to take steps to ease access to government-wide acquisition solutions for AI services and products, potentially including the creation of an AI acquisition resource guide.
Increasing AI Talent in Government
- Under the EO, federal agencies are broadly directed to hire and retain AI talent. The EO establishes more-specific plans for bolstering the federal AI workforce and making use of AI talent: the Office of Science and Technology Policy (OSTP) and OMB will establish AI hiring priorities and plans; the Assistant to the President and Deputy Chief of Staff for Policy will convene an AI and Technology Talent Task Force; and the Secretary of Defense will prepare a report on gaps in AI talent for national defense. AI training and familiarization programs will also be made available by the heads of each agency for employees in AI-relevant roles at all levels. As we reported here, Congress has already made progress addressing the need for AI talent in the federal government, which includes two pending bills, S.1564 – AI Leadership Training Act and S.2293 – AI LEAD Act.
Section 11. Strengthening American Leadership Abroad
Section 11 of the EO outlines a set of actions to bolster U.S. leadership in the global effort to harness the benefits of AI while responding to its challenges. The EO designates the Departments of State, Commerce, Homeland Security, and Energy with the primary authority to engage with international allies and partners, in collaboration with other relevant U.S. agencies, notably including the U.S. Agency for International Development. These efforts will seek to foster greater understanding of U.S. policies related to AI, encourage the responsible global development of the technology, and promote international collaboration.
Section 11 directs the aforementioned U.S. agencies to complete the following actions, which collectively aim to fortify U.S. global leadership in the development of AI and promote the safe, responsible, and interoperable deployment of the transformative technology:
- Establish a comprehensive international framework to manage both the risks and benefits of AI and encourage allies to make voluntary commitments similar to those by U.S. companies—likely a nod to recent commitments made by private companies, including Amazon, Google, Microsoft, and OpenAI and last week’s announcement by the G7 on a voluntary AI code of conduct.
- Develop a plan for global engagement on AI standards, covering topics such as AI terminology, best data practices, trustworthiness of AI systems, and AI risk management. These efforts should adhere to the principles in the AI Risk Management Framework and the National Standards Strategy for Critical and Emerging Technology.
- Create an AI in Global Development Playbook, incorporating the principles of the AI Risk Management Framework and applying them to various international contexts. Doing so will likely require close coordination with international partners to ensure appropriate “translations” to local contexts that may differ from those in the United States.
- Establish a Global AI Research Agenda to guide the objectives and execution of AI-related research beyond the United States. The Agenda will address safety, responsibility, benefits, sustainability, and labor-market implications of AI adoption in different international contexts.
- Enhance cooperation with international allies and partners in preventing, responding to, and recovering from potential disruptions to critical infrastructure as a result of AI incorporation or malicious AI use. The Secretary of DHS has authority over the adoption of AI safety and security guidelines for critical infrastructure.
Overall, this new EO incorporates a sweeping set of provisions tailored to the evolving AI landscape. It is clear throughout each section that the government is primarily focused on the safe and secure development and deployment of AI. The EO seeks to maximize the potential benefits of AI while also addressing rising concerns about its potential harms, building on the Administration’s efforts to strike a delicate balance between innovative and protectionary action.
Crowell & Moring LLP and Crowell & Moring International continue to monitor congressional and executive branch efforts to regulate AI. Our lawyers and public policy professionals are available to advise any clients who want to play an active role in the policy debates taking place right now or who are seeking to navigate AI-related concerns in financial services, intellectual property, privacy, health care, government contracts, and other areas.