AI Safety and Security (E.O. 14110)

Summary

Creates rules for the U.S. government and companies to make sure Artificial Intelligence is safe, protects people's rights, and helps the country.

What problem does this solve?

Artificial intelligence is growing quickly and could be used in harmful ways, such as spreading false information, taking jobs, or threatening national security. This order sets up a government-wide plan to manage AI's risks by creating safety rules, protecting workers, and ensuring fairness for everyone.

What does this order do?

Requires safety reporting for powerful AI models

Forces companies developing the most powerful AI systems to report their safety test results and other key information to the federal government to ensure they are not a threat.

Establishes standards for AI safety and security

Directs the National Institute of Standards and Technology (NIST) to create guidelines and best practices for developing and using AI safely, including ways to test for risks.

Creates Chief AI Officer role in federal agencies

Requires every major federal agency to designate a Chief Artificial Intelligence Officer. This person will be responsible for coordinating the agency's use of AI and managing its risks.

Protects against AI-driven discrimination

Directs federal agencies to prevent and address unlawful discrimination caused by AI in areas like housing, government benefits, and the criminal justice system.

Develops tools to identify AI-generated content

Orders the creation of standards for watermarking and detecting AI-generated content. This helps people know when images, videos, or text were made by AI.

Supports workers affected by AI

Requires the Secretary of Labor to create principles and best practices for employers. This helps protect workers from job loss, unfair monitoring, and other harms from AI in the workplace.

Streamlines immigration for AI experts

Makes it easier for foreign nationals with skills in AI and other new technologies to get visas to study, work, or conduct research in the United States.

Manages AI risks in critical infrastructure

Requires government agencies to assess how AI could make critical systems like the power grid, transportation, and financial services vulnerable to attacks or failures.

Promotes AI innovation and competition

Launches programs to give researchers access to AI tools and data. It also supports small businesses to ensure that large companies do not control the AI market.

Establishes a White House AI Council

Creates a new council within the White House to coordinate AI policy and make sure all federal agencies are working together effectively on AI issues.

Who does this affect?

Technology companies developing AI
Federal government agencies
American workers and consumers

What is the real world impact?

•

Protects against discrimination and bias

Prevents AI tools from being used in ways that are unfair to people, especially in areas like hiring, housing, and the justice system. It requires checks for bias and discrimination.

•

Addresses public fears about AI risks

Responds to growing public concern about AI's potential to cause job loss, spread misinformation, and create deepfakes. It creates rules for content labeling and worker protection.

•

Balances innovation with national security

Encourages AI development to boost the economy and solve big problems. It also puts strict controls on powerful AI models to prevent them from being used for cyberattacks or creating weapons.

•

Establishes U.S. leadership in global AI governance

Aims to position the United States as the global leader in AI policy. It sets international standards for safety, security, and ethics to guide how other nations develop and use the technology.

When does this start?

This order sets multiple deadlines for different government agencies to complete specific tasks related to AI safety, innovation, and governance.

Require safety reporting from AI developers

Within 90 days (by January 28, 2024), the Secretary of Commerce must require companies developing powerful AI models to report their safety testing results to the government.

Streamline visas for AI talent

Within 90 days (by January 28, 2024), the Secretaries of State and Homeland Security must take steps to speed up visa processing for experts in AI and other critical technologies.

Issue guidance for federal AI use

Within 150 days (by March 28, 2024), the Director of the Office of Management and Budget must issue guidance for federal agencies on how to use AI safely and effectively.

Develop principles for workers

Within 180 days (by April 27, 2024), the Secretary of Labor must develop and publish principles and best practices for employers to protect workers from the potential harms of AI.

Create standards for AI-generated content

Within 240 days (by June 26, 2024), the Secretary of Commerce must identify standards and tools for authenticating content and labeling AI-generated material.

Establish guidelines for AI safety

Within 270 days (by July 26, 2024), the National Institute of Standards and Technology (NIST) must establish guidelines and best practices for developing safe, secure, and trustworthy AI systems.

Report on AI in the criminal justice system

Within 365 days (by October 30, 2024), the Attorney General must submit a report on the use of AI in the criminal justice system, including in sentencing, policing, and risk assessments.

Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence