On Oct. 30, 2023, the president issued an Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence. The 117-page order includes a range of directives, many of which are aimed at promoting the domestic development of AI technologies while also addressing the perceived risks. The order lays the groundwork for further action by requiring federal executive agencies to conduct research studies and implement other measures to prepare for the further proliferation of AI technologies.

The order follows prior government pronouncements concerning the recent proliferation of generative AI technologies. In October 2022, the Office of Science and Technology Policy published the “Blueprint for an AI Bill of Rights”, and earlier this year the National Institute of Standards and Technology released the “AI Risk Management Framework.” Furthermore, this past July, the White House announced that Amazon, Anthropic, Google, Inflection, Meta, Microsoft and OpenAI had provided the administration certain voluntary commitments connected to their development of AI technologies. In many respects, this order codifies and builds upon these prior actions.

Consumer protection is a driving force behind many of the directives in the order. For example, to prevent fraud, the order directs the Secretary of Commerce to submit a report identifying standards, tools, methods and practices for labeling and/or authenticating AI-generated content (e.g., via watermarks). While this report does not currently impose binding requirements, the future emergence of labeling standards could significantly impact the private-sector adoption of AI technology, as many businesses fear customers may react negatively to the use of AI-generated content. The order also addresses consumer discrimination and bias, both of which were key themes of the earlier “Blueprint for an AI Bill of Rights.”

The order also aims to promote the safety and security of AI technology. Notably, the order requires that the largest AI developers disclose certain security-related information with the federal government on an ongoing basis. This includes, but is not limited to, information about the physical and digital cybersecurity measures in place, employee cybersecurity training and the results of “red-team” security testing. Unlike the other directives in the order, the president cites his powers under the Defense Production Act (DPA) of 1950 to enact these disclosure requirements. Relying on this legal authority is a noteworthy development, as the DPA is intended for use in relation to national defense matters. In practice, these reporting requirements may apply only to a limited subset of AI technology developers, all of whom have agreed to such disclosures as a component of the aforementioned voluntary commitments last July.

Much of the lengthy order is dedicated to instructing specific federal agencies to analyze the impacts of AI technology on the public and private sector activities they govern. For example, under the order, the Secretary of Labor is instructed to prepare and submit “a report analyzing the abilities of agencies to support workers displaced by the adoption of AI and other technological advancements.” The order also directs the Secretary of Education to develop resources addressing the “safe, responsible, and nondiscriminatory uses of AI in education.” There are numerous other examples of similar directives aimed at other federal agencies, most of which are tailored to address specific benefits or risks currently associated with AI technology. Agencies are also required to designate permanent “Chief AI Officers.” These officers will participate in an interagency council tasked with “coordinating the development and use of AI in agencies’ programs and operations. . .“

The order also includes directives aimed at inducing the development of AI technologies domestically. This includes a directive under which the Secretary of State must establish a program “to identify and attract top talent in AI . . . including overseas educational components to inform top STEM talent of nonimmigrant and immigrant visa options.” Under the order, the Secretary of Homeland Security is also instructed to “review and initiate any policy changes the secretary determines necessary and appropriate to clarify and modernize immigration pathways for experts in AI and other critical and emerging technologies.”

While the number of enforceable requirements is quite limited, this broad-reaching order may be remembered most as a harbinger of future federal AI requirements. In the remarks accompanying the release of the order, the President stressed the need for federal legislation governing AI technology. In the absence of such legislation, the regulatory authority available to the current and future administrations will remain limited.