An Examination Of Certain Key Features Of The New White House Executive Order On Artificial Intelligence

Abstract

This article examines certain key features of the White House Executive Order on Artificial Intelligence issued on October 30, 2023. It places this Executive Order in the context of prior US governmental actions related to AI and discusses the legal strength of the Executive Order. The article takes the position that as much of the Executive Order consists of instructions to government agencies to develop guidance and draft regulations for further consideration the true impact of much of the Executive Order is yet to be seen. The article states that the most immediate direct impact on the private sector will be the Executive Order’s private sector reporting requirements related to the development of dual-use foundation models and large scale computing clusters which pose risks to national security. The article states that the US National Institute of Standards and Technology will likely continue to play a central role in the evolution of AI regulatory initiatives and may continue to look to international AI related ISO and IEC standards when doing so, that red-teaming precedents from cybersecurity practices will likely be used to vet AI systems and that Know Your Customer requirements drawn from the example of the financial services sector may well be put in place for US based cloud service providers. The article states that given the content of the Executive Order and the inherent weaknesses in Executive Orders more generally it is unlikely that the US Congress’ focus on potential AI legislation will cease.