President Biden has issued an Executive Order focused on the responsible development and application of artificial intelligence (AI). This move is part of a broader strategy to manage AI-related risks while encouraging innovation, and is the first-ever executive order of its kind from the U.S. government.
The executive order is made up of eight broad areas.
New standards for AI safety and security
The Executive Order outlines extensive measures to enhance AI safety and security.
- Mandatory safety reporting: Developers of high-risk AI systems must share safety tests and other key information with the U.S. government. This is in line with the Defense Production Act.
- AI safety standards: The National Institute of Standards and Technology will set standards for red-team testing to ensure AI safety. The Department of Homeland Security will enforce these standards across critical infrastructure sectors, and establish the AI Safety and Security Board. The Departments of Energy and Homeland Security will start addressing AI systems’ threats to critical infrastructure, as well as chemical, biological, radiological, nuclear, and cybersecurity risks.
- Biological risks: Agencies funding life-science research will set new screening standards for biological synthesis, mitigating risks exacerbated by AI.
- Combatting AI-enabled fraud: The Department of Commerce will develop guidance for content authentication and watermarking to clearly label AI-generated content.
- Cybersecurity initiative: An advanced cybersecurity program will leverage AI to identify and fix vulnerabilities in essential software.
- National security memorandum: Directed by the National Security Council and the White House Chief of Staff, this document will set further AI-related actions for military and intelligence use, and to counter adversarial use of AI.
Protecting Americans’ privacy
To address the privacy risks heightened by AI, President Biden has called for bipartisan data privacy legislation and outlined key actions.
- Privacy-preserving AI development: A federal initiative will fast-track the creation of advanced AI techniques that protect individual data during system training.
- NSF-funded research coordination network: The National Science Foundation will fund a network focused on innovative privacy technologies, including cryptographic tools. Collaboration with this network will help federal agencies implement these technologies.
- Review of data collection practices: Federal agencies will scrutinize their data sourcing methods, particularly data acquired from brokers. Revised guidelines will address potential AI-related privacy risks.
- Evaluation guidelines: New standards will be established to help federal agencies assess the efficacy of privacy-preserving techniques, especially those used in AI systems.
Advancing equity and civil rights
The Order provides directives to combat algorithmic discrimination.
- AI and discrimination mitigation: Directives will be issued to landlords, Federal benefits programs, and federal contractors to prevent AI algorithms from amplifying discrimination.
- Algorithmic discrimination accountability: Training, technical support, and coordination between the Department of Justice and Federal civil rights offices will establish best practices for investigating and prosecuting AI-related civil rights violations.
- AI in criminal justice fairness: Guidelines will be developed to ensure equitable AI use in various criminal justice applications, including sentencing, parole and probation, pretrial activities, risk assessments, surveillance, and predictive policing.
Standing up for consumers, patients, students
The Executive Order aims to protect consumers from the risks posed by AI. It advocates for the responsible use of AI in healthcare and education, and calls for reports on harmful practices.
- AI in Healthcare: The Department of Health and Human Services will focus on promoting responsible AI use in healthcare and drug development. A safety program will be established to address reports of AI-related harms or unsafe practices in healthcare.
- AI in Education: Resources will be created to assist educators in implementing AI-powered educational tools, including personalized tutoring, to transform the learning experience.
The order outlines best practices to address job displacement, workplace bias, and data collection.
- Guiding principles: A set of principles and best practices will be developed to tackle issues related to job displacement, labor standards, and workplace equity. These guidelines aim to prevent employers from under-compensating workers, conducting unfair job evaluations, or undermining workers’ organizational efforts.
- AI’s impact on labor market: A comprehensive report will be produced to study the potential effects of AI on the labor market. The report will also explore options for bolstering federal support for workers who are impacted by labor disruptions, including those caused by AI technologies.
Promoting innovation and competition
The Order aims to maintain American leadership in AI. It also calls for actions to support small developers and entrepreneurs.
- National AI research: Launch a pilot program that offers AI researchers and students access to essential AI resources and data. This will catalyze research efforts in critical sectors such as healthcare and climate change through expanded grants.
- Support for small developers: Create a competitive AI landscape by offering technical assistance and resources to small developers and entrepreneurs. This initiative will aid in the commercialization of AI breakthroughs and enable the Federal Trade Commission to exercise its regulatory powers.
- Streamlining immigration for AI experts: Utilize existing regulations to ease the process for highly skilled immigrants and nonimmigrants with expertise in critical sectors to study, stay, and work in the United States. This involves modernizing and streamlining visa criteria, interviews, and reviews.
Advancing American leadership abroad
The administration plans to collaborate with international partners to develop global AI governance frameworks.
- International AI collaboration: The State Department and Commerce Department will spearhead efforts to form comprehensive international frameworks focused on maximizing AI benefits and mitigating risks.
- Global AI standards: Fast-track the development and adoption of essential AI safety standards in partnership with international entities, ensuring that AI technology is secure, trustworthy, and interoperable.
- Global AI responsibility: Champion the deployment of AI solutions abroad that are not only safe and responsible but also aligned with human rights, aimed at addressing global challenges like sustainable development and infrastructure security.
Ensuring responsible and effective government use of AI
The Executive Order issues guidance for federal agencies’ AI use. It encourages rapid hiring of AI professionals and outlines plans for government-wide AI training.
- AI deployment guidelines for federal agencies: Issue comprehensive guidance that sets clear standards for AI usage within agencies, focusing on rights, safety, and efficient procurement.
- Streamlined AI procurement: Facilitate quicker, more cost-effective acquisition of AI products and services through optimized contracting processes.
- AI talent surge: Under the coordination of multiple offices including the Office of Personnel Management and U.S. Digital Service, accelerate the hiring of AI professionals. This will be complemented by specialized AI training for government employees across various departments.
An Executive Order is a directive issued by the President of the United States to manage the operations of the federal government. It carries the force of law but is not legislation; rather, it is meant to implement or interpret a law that already exists, or to specify a course of action in the management of the federal government. Executive Orders are generally issued to federal agencies and officials and can dictate how they should operate or what actions they should take.
You can find the text of Biden’s executive order on AI here.
[Image courtesy: The White House]