Impending Government Mandate for AI Companies to Report Foundation Model Development

In a significant move by the US government, companies involved in the development of foundational AI models, such as OpenAI's GPT-4 and Google's Gemini, will soon be required to inform federal authorities about their advancements. This development comes in the wake of comments made by US Secretary of Commerce Gina Raimondo at a Stanford University's Hoover Institute event, emphasizing the use of the Defense Production Act to survey and review large language model training and safety data. Secretary Raimondo's remarks highlight the government's initiative to gather detailed information whenever a new large language model is trained. The focus is on understanding the safety aspects of these models, ensuring they align with national security and public safety standards. This move is a part of President Biden's comprehensive AI executive order, which emphasizes the need for companies developing foundation models posing significant risks to national security, economy, or public health to notify the government and share safety test results. While current models like GPT-4 might not meet the threshold for government oversight due to computing power limitations, future models with unparalleled computing capabilities present a potential national security risk. The Defense Production Act, previously invoked in 2021 for pandemic-related purposes, now plays a crucial role in regulating these advanced AI technologies. Another critical aspect of the executive order addresses the usage of US cloud services by foreign entities. Companies like Amazon, Google, and Microsoft will be required to report any non-US usage of their cloud services for training large language models, as stated by Raimondo. This is a step towards enhancing transparency in the global utilization of American cloud computing resources. The exact timeline for these new requirements to take effect remains undisclosed, with an imminent announcement anticipated given the looming deadline. This policy marks a pivotal moment in the regulation and oversight of AI development, reflecting the government's proactive stance in ensuring the safe and secure advancement of artificial intelligence.