WORLD ROUNDUPNuclear Arms Control Is for Realists | Understanding the Deterrence Gap in the Taiwan Strait | Milei’s Swing into Normality Might Not Last, and more

The U.S. Plans to ‘Lead the Way’ on Global AI Policy  (Alan Charles Raul and Alexandra Mushka, Lawfare)
Policymakers around the world took significant steps toward regulating artificial intelligence (AI) in 2023. Spurred by the launch of revolutionary large language models such as OpenAI’s GPT series of models, debates surrounding the benefits and risks of AI have been brought into the foreground of political thought. Indeed, over the past year, legislative forums, editorial pages, and social media platforms were dominated by AI discourse. And two global races have kicked into high gear: Who will develop and deploy the most cutting-edge, possibly risky AI models, and who will govern them? 
In the wake of this competition, it is worth examining whether the United States will yield policy primacy on AI to Europe, or others, as it largely has done in the field of data privacy—or whether it will instead assert leadership on digital governance commensurate with its lead in the digital technology itself. The plethora of federal initiatives adopted in response to the deployment of capable AI systems with significant computational power supports the latter thesis: The United States intends to run ahead of the field on AI governance, analogous to U.S. leadership on cybersecurity rules and governance—and unlike the policy void on privacy that the federal government has allowed the EU to fill. Various policy developments discussed below support this conclusion, chief among them the fact that the aggressive timeline of government action imposed by President Biden’s October 2023 AI executive order means the requirements, imperatives, and guidelines that order sets into motion will almost certainly be in force before the EU’s provisional AI Act is adopted and implemented. Indeed, the White House recently announced that every 90-day deadline set forth by the order has been met. Notably, “developers of the most powerful AI systems” are already required “to report vital information” to the Department of Commerce, including the results of safety testing. Nine agencies have submitted risk assessments to the Department of Homeland Security regarding the use of AI in critical infrastructure. The intense level of agency engagement called for by President Biden led the order to be viewed by some in Congress and industry as too powerful, triggering a “campaign to take [it] down,” or “defang the industry-facing sections.”