US President Joe Biden signed an order to regulate generative AI and protect the public. The Biden administration seeks to increase safety around artificial intelligence (AI) and protect consumers and workers from AI-related risks.
The order sets eight objectives for AI development – safety, security, privacy, fairness, consumer protection, worker support, innovation, and responsible government use.
These initiatives will be implemented in stages over the next year. Minor safety improvements will be put in place in the next 90 days, while comprehensive reporting mandates may take up to a year. An “AI council” led by White House Deputy Chief of Staff Bruce Reed will oversee these initiatives.
US government agencies are tasked to create standards to protect against AI misuse in biological engineering, establish content authentication best practices, and build advanced cybersecurity. The National Institute of Standards and Safety (NIST) will create standards to test AI models before public release, while the Department of Energy and Department of Homeland Security will address potential infrastructure threats and cybersecurity risks. Developers of large AI models must share safety test results.
Defense Production Act to ensure risky AI models don’t get in the hands of the public
The Biden administration has set new standards for AI safety and security, including reporting requirements for developers whose models might impact national or economic security. The Defense Production Act will ensure companies developing risky models must notify the federal government and share safety test results before releasing them to the public.
The EO will establish a cybersecurity program based on the AI Cyber Challenge to develop AI tools that can autonomously root out security vulnerabilities in software infrastructure.
Federal agencies to develop watermarking system to verify AI-generated content
The Department of Commerce will develop a watermarking system for federal agencies to verify whether content is AI-generated. The government endorses technical standards to lead to wider adoption of digital watermarking. This will help federal officials ensure reliable government messages.
AI companies like OpenAI, Alphabet, and Meta Platforms agreed to watermark AI-generated content voluntarily to increase safety.
AI Bill of Rights extends to include civil rights violations
To protect user privacy, the White House urged Congress to pass data privacy regulations and support the development of “privacy-preserving” techniques.
Last October, the White House released an AI Bill of Rights that aims to combat discriminatory AI practices. The new Executive Order extends guidance to landlords, federal programs, and contractors, and directs the Department of Justice to develop best practices for investigating and prosecuting AI-related civil rights violations. The EO also prioritises the development of privacy-preserving techniques, establishes a safety program for AI-based medical practices, and supports educators in using AI-based educational tools.
Job security against AI
The Biden administration admits that AI is a boon for business, but it also threatens job security. The administration will provide guidelines and best practices for employers to follow. At the same time, the Department of Labor and the Council of Economic Advisors study how AI could affect the job market. The administration also recognizes the importance of responsible government development and deployment of AI.
The US government is launching AI.gov, a new jobs portal, on Monday to attract more AI talent to federal jobs. The portal will provide information and guidance on available fellowship programs.
The White House Deputy Chief of Staff, Bruce Reed, has referred to the order as the most robust set of actions any government has taken to ensure AI security. The order also addresses privacy, housing discrimination, and job displacement. Reed stated that it is the next step in an aggressive strategy to leverage the benefits of AI while mitigating the risks involved.
The order sets eight objectives for AI development – safety, security, privacy, fairness, consumer protection, worker support, innovation, and responsible government use.
These initiatives will be implemented in stages over the next year. Minor safety improvements will be put in place in the next 90 days, while comprehensive reporting mandates may take up to a year. An “AI council” led by White House Deputy Chief of Staff Bruce Reed will oversee these initiatives.
US government agencies are tasked to create standards to protect against AI misuse in biological engineering, establish content authentication best practices, and build advanced cybersecurity. The National Institute of Standards and Safety (NIST) will create standards to test AI models before public release, while the Department of Energy and Department of Homeland Security will address potential infrastructure threats and cybersecurity risks. Developers of large AI models must share safety test results.
Defense Production Act to ensure risky AI models don’t get in the hands of the public
The Biden administration has set new standards for AI safety and security, including reporting requirements for developers whose models might impact national or economic security. The Defense Production Act will ensure companies developing risky models must notify the federal government and share safety test results before releasing them to the public.
Expand
Federal agencies to develop watermarking system to verify AI-generated content
The Department of Commerce will develop a watermarking system for federal agencies to verify whether content is AI-generated. The government endorses technical standards to lead to wider adoption of digital watermarking. This will help federal officials ensure reliable government messages.
AI companies like OpenAI, Alphabet, and Meta Platforms agreed to watermark AI-generated content voluntarily to increase safety.
AI Bill of Rights extends to include civil rights violations
To protect user privacy, the White House urged Congress to pass data privacy regulations and support the development of “privacy-preserving” techniques.
Last October, the White House released an AI Bill of Rights that aims to combat discriminatory AI practices. The new Executive Order extends guidance to landlords, federal programs, and contractors, and directs the Department of Justice to develop best practices for investigating and prosecuting AI-related civil rights violations. The EO also prioritises the development of privacy-preserving techniques, establishes a safety program for AI-based medical practices, and supports educators in using AI-based educational tools.
Job security against AI
The Biden administration admits that AI is a boon for business, but it also threatens job security. The administration will provide guidelines and best practices for employers to follow. At the same time, the Department of Labor and the Council of Economic Advisors study how AI could affect the job market. The administration also recognizes the importance of responsible government development and deployment of AI.
The US government is launching AI.gov, a new jobs portal, on Monday to attract more AI talent to federal jobs. The portal will provide information and guidance on available fellowship programs.
The White House Deputy Chief of Staff, Bruce Reed, has referred to the order as the most robust set of actions any government has taken to ensure AI security. The order also addresses privacy, housing discrimination, and job displacement. Reed stated that it is the next step in an aggressive strategy to leverage the benefits of AI while mitigating the risks involved.
