Access Now welcomes the release of President Biden’s highly anticipated Executive Order (EO) on Safe, Secure, and Trustworthy Artificial Intelligence (AI), which will strengthen the regulation of U.S. federal agencies’ use of AI. The order underscores the need to address the multifaceted societal impacts of AI and reinforces the importance of international cooperation in tackling harm caused to human rights. To ensure the success of the EO, however, Congress must pass its principles into law and strictly regulate high-risk AI systems impacting vulnerable communities, including redline prohibitions on certain biometric recognition technologies.
The Executive Order represents the culmination of the administration’s yearlong initiative to address the challenges and opportunities of AI, following the release of the White House Blueprint for an AI Bill of Rights (AI Bill of Rights) and a set of 15 voluntary commitments by leading companies. Today’s executive order serves as a catalyst for new safety standards for AI, advancing civil rights and equity, supporting the workforce, advancing American leadership abroad, addressing algorithmic discrimination, and ensuring the responsible and effective use of AI within the government.
One of the most notable aspects of EO is its call on Congress to pass comprehensive data privacy legislation. The EO goes on to leverage the U.S. government’s influence as a major consumer of AI technologies to ensure federal agencies using these technologies do no harm. The order mandates assessments of AI models before federal agencies use them. It invokes the Defense Production Act, requiring developers of powerful AI systems to share safety test results and critical information with the U.S. government. Furthermore, the EO also directs the establishment of an advanced cybersecurity program to develop AI tools to find and fix vulnerabilities in critical software.
Moreover, the EO directs the National Institute of Standards and Technology to set standards for extensive red-team testing — and the Department of Commerce to develop guidance for content authentication and watermarking to label AI-generated content. In a move to strengthen the workforce, the order facilitates immigration opportunities for highly skilled workers. Additionally, federal agencies will need to assess the size of the AI workforce to support workforce growth and development.
In September, Access Now joined over 65 civil, human, and digital rights organizations urging the Biden-Harris administration to take decisive action and make the AI Bill of Rights binding U.S. government policy for the federal government’s use of these systems in the AI EO. Additionally, Access Now has engaged with Senate offices working on an AI legislative framework and joined partners in urging greater engagement with civil society.